Delaware courts largely have privatized enforcement of fiduciary duties in public corporations and have expressly acknowledged this judicial policy. The Delaware courts also recognize that so encouraging private enforcement creates an obvious danger: Plaintiffs' attorneys, especially in class actions where there is no strongly interested plaintiff, may make litigation-related decisions primarily with a view to advancing their own economic interests, rather than advancing the interests of the corporation or shareholders that they purport to represent. Such decisions have the potential to impose substantial, litigation-related agency costs on corporations, shareholders, and the courts, if not appropriately curbed through judicial monitoring of settlements and fee awards.
This Article examines the development of Delaware law with respect to merger-related class actions, which have become the dominant form of shareholder litigation in Delaware. We offer two broad alternative hypotheses as to what drives merger-related class actions in Delaware: a "shareholder champion" hypothesis and a "self-interested litigator" hypothesis. We then examine intensively all large mergers in 1999-2001 where the target was a publicly traded Delaware company and all class actions filed with respect to those mergers. We conduct statistical analyses as well as a detailed qualitative analysis of the 104 class actions filed during those years.
The pattern that we observe is redolent of a pattern of opportunistic filings, of a lawyer-driven process rather than a true client-driven process: systematic behavior with respect to which mergers were challenged; early and frequent complaints filed; a very high percentage of dismissed cases never reached a judgment on the merits; the absence of a single case that has been decided in favor of the plaintiffs on the merits; settlements tending to reflect free riding by plaintiffs' attorneys; plaintiffs' attorneys failing to challenge special negotiating committees' decisions or competing offers; attorneys with "real" clients and from outside the "traditional" Delaware plaintiffs' bar who were far more vigorous in their litigation efforts; no settlements overturned by the Delaware courts; plaintiffs' attorneys' fee awards in settlements usually paid by defendants and not out of common funds, and largely unchallenged; and plaintiffs' attorneys' fees representing a strikingly low percentage of claimed recoveries (but attractive on an hourly basis), which may well indicate that the attorneys added little value to the recoveries.
We then offer suggestions as to changes in pleading standards and the Delaware courts' approach to reviewing settlements and plaintiffs' attorneys' fees that would help curb the excesses of class action litigation without seriously undermining the constructive role that plaintiffs' attorneys have the potential to play in policing corporate misgovernance with respect to mergers.
Delaware courts have largely privatized enforcement of fiduciary duties in public corporations. In In re Fuqua Industries, Inc. Shareholder Litigation,1 Chancellor Chandler expressly acknowledged this judicial policy. He noted that Delaware courts implement it partly by allowing private attorneys, working on a contingent fee basis, to initiate and maintain derivative and class actions in the names of "nominal shareholder plaintiffs."2 Attorneys are subject only to the relatively weak constraints that they must inform their "clients" and receive their consent before they file shareholder suits. Further, Delaware courts use cost and fee shifting mechanisms to "economically incentivize"3 those attorneys to initiate such suits.4
Chancellor Chandler also explained that Delaware courts have adopted this policy because they believe that the plaintiffs' bar is capable of performing a valuable "service on behalf of shareholders."5 Plaintiffs' attorneys understand "abstruse issues of corporate governance and fiduciary duties"6 far better than do most shareholders. Consequently, they are uniquely qualified to identify situations in which principles of corporate governance have been violated or fiduciary duties have been breached and then to initiate lawsuits seeking corrective action.7
At the same time, Delaware courts have recognized that encouraging private enforcement creates an obvious danger. Plaintiffs' attorneys may make litigation-related decisions primarily to advance their own economic interests rather than those of the corporations or shareholders that they purport to represent.8 Such decisions have the potential to impose substantial, litigation-related agency costs on corporations, shareholders, and the courts.9
Concerns about possible litigation-related agency costs have led Delaware's courts to impose two major constraints on shareholder derivative suits. They have interpreted with considerable rigor the long-standing requirement that a derivative plaintiff must either make demand on the defendant corporation's board of directors or plead with particularity why demand would be futile.10 In addition, they have held that a special litigation committee, appointed by a defendant corporation's board of directors, has the right to investigate plaintiff's claims and to seek dismissal of a derivative suit, even where plaintiff has established that demand is excused.11 One explicit purpose of both these requirements is to weed out opportunistic claims -often referred to as "strike suits"-that may generate substantial fee awards without producing meaningful benefits for the defendant corporation or its shareholders.12
Neither Delaware courts nor Delaware's legislature has imposed comparable constraints on shareholder class actions, the other principal form of representative shareholder litigation.13 Yet, as Professors Thompson and Thomas point out in their innovative study of lawsuits filed in Delaware Chancery Court in the years 1999-2000, class actions and, in particular, acquisition-related class actions appear to have become "the dominant form of corporate litigation, outnumbering derivative suits by a wide margin."14 Whether the apparent preference of the plaintiffs' bar for class action arises from a lack of constraints on opportunistic litigation, from a proliferation of wrongdoing in connection with mergers and acquisitions, or from both, is unclear.
At the time we first became aware of Thompson and Thomas's findings, we were in the early stages of a study of one category of merger-related class actions that we believed was particularly vulnerable to opportunistic behavior by the plaintiffs' bar-suits challenging so-called "sales of control."15 We had become interested in this area as a consequence of the experience of one of us in connection with two class actions challenging the fairness of transactions, seemingly negotiated at arm's length, in which Delaware corporations were sold for cash. Plaintiffs' attorneys filed complaints in both suits almost immediately after the challenged transactions were announced. In neither suit, however, did plaintiffs' attorneys find any evidence to support their claims that the corporation had been sold for less than the highest value reasonably available. As a result, in neither case did they obtain any monetary recovery for the plaintiff class. Instead, both cases settled because defendants agreed to disclose additional information that, in both cases, tended to bolster defendants' contention that they had sought and obtained the highest value reasonably available for the subject corporation's stockholders. However, despite the absence of any evidence of wrongdoing, defendants or their successors in both cases agreed not only to settle but also to pay plaintiffs' attorneys' fees and expenses, up to agreed-upon amounts, if awarded by the court.
As a member of the plaintiff class in one of these suits, Steiner v. CalMat Co.,16 and as counsel to one of his colleagues who was a member of the plaintiff class in the second, In re Banctec, Inc. Shareholders Litigation,17 Professor Weiss prepared objections to both settlements. In both cases, the objections focused on plaintiffs' attorneys' fee request. Both objectors argued that no fees should be awarded because plaintiffs' attorneys initially filed complaints without having any reasonable factual basis for the claims they had made, that plaintiffs had found no evidence of wrongdoing, and that the settlements in question provided no meaningful benefits to the plaintiff class.18
Although both objections achieved substantial success,19 what seemed more significant to us was that Delaware law appeared to make it attractive for plaintiffs' attorneys to file class actions challenging "sale-of-control" transactions whether or not fiduciary breaches appeared to have occurred. A complaint that simply asserted that a corporation's board had approved a "sale-of-control" and that then alleged in general terms that the directors had not obtained "the best value reasonably available" could not be dismissed because, under Delaware law, such allegations sufficed to shift to defendant directors the burden of proving that they had not violated their Revlon duties.20 Moreover, compendia of unreported decisions that plaintiffs' attorneys in both cases filed in support of their fee requests suggested to us that many other class actions had been filed on the basis of similar generalized allegations of wrongdoing, had also been settled for nonmonetary relief and-in the absence of an objection by a "gadfly" such as Professor Weiss-had resulted in substantial attorneys' fee awards.21 In effect, we began to suspect that, at least with respect to sale-of-control transactions, Delaware law created a sort of safe harbor for strike suits. That is, Delaware law made it too easy for plaintiffs' attorneys, in connection with any sale-of-control transaction, to initiate a class action that would have significant settlement value. Plaintiffs' attorneys could threaten the timing of the transaction and impose on the board of the selling corporation the litigation costs involved in proving that it had proceeded in accord with its duties under Revlon.22
Examination of a preliminary draft of Thompson and Thomas's paper led us to broaden the scope of our study. Thompson and Thomas found that most of the class actions filed in 1999-2000 were directed at mergers that involved some conflict of interest-a squeeze out by a controlling shareholder, a management buy out (MBO), or a sale of control to a third party in connection with which the selling corporation's managers allegedly diverted a disproportionate share of the proceeds to themselves. Relatively few class actions challenged sales of control in the absence of self-dealing. We had already hypothesized that plaintiffs' attorneys' decisions concerning whether to file and how to prosecute merger-related class actions resulted primarily from Delaware law-in particular, that the pleading rules, burden of proof, and standard of review made it easy to file a claim that would survive a motion to dismiss and provide plaintiffs with economic leverage to press for a settlement and a fee award, regardless of whether fiduciary duties had been breached. Thompson and Thomas's data led us to hypothesize further that the presence of a conflict of interest in connection with a merger would make it more attractive for plaintiffs' attorneys to file a class action because unfair self-dealing was more likely to have occurred. In our view, that potential also made it more likely that class actions challenging mergers involving conflicts of interest had produced substantial benefits for members of the plaintiff class as well as the attorneys who filed them.
We further revised our hypotheses after we began to examine litigation documents. They revealed that during the three-year study period, boards of target companies in mergers involving conflicts of interest routinely appointed special negotiating committees ("SNCs") composed of independent directors and charged those committees with responsibility for negotiating the best possible terms on behalf of the company's public shareholders.24 Plaintiffs' attorneys, however, routinely filed class actions challenging such mergers as soon as they were announced, without regard to whether the target company's board had appointed an SNC or agreed to final terms. Moreover, if plaintiffs file a suit anticipating a target's board will breach its fiduciary duties, and if an SNC subsequently negotiates some improvement in the terms of a proposed merger (as often is the case), Delaware law presumes that the improvement is attributable in whole or in part to plaintiffs' attorneys' efforts.25 Consequently, even if plaintiffs' attorneys accept as fair the merger terms an SNC has negotiated, Delaware law allows them to seek compensation for the benefits that their lawsuit is presumed to have generated.26
This combination of corporate practice and Delaware law provides plaintiffs' attorneys with substantial leverage in mergers involving conflicts of interest. Unless defendants are confident both that they can prove a negative-that plaintiffs' attorneys' efforts had no impact on any improved terms negotiated by an SNC-and are further prepared to incur the litigation costs involved in defending a suit seeking attorneys' fees, negotiating an agreement to pay plaintiffs' attorneys' fees probably will appear to present an attractive alternative. Thus, in most mergers involving conflicts of interest and especially in mergers involving sales of control, the combination of corporate practice and Delaware law appears to have provided plaintiffs' attorneys with substantial incentives to file class actions, regardless of whether it appeared that fiduciary duties had been or would be breached.27
We also recognized, though, that the mere presence of these incentives to pursue opportunistic litigation did not necessarily mean that plaintiffs' attorneys had been acting as "unfaithful champions"28 of shareholders' interests. After analyzing all "acquisition-related class actions"29 in Delaware Chancery Court in 1999-2000 from a somewhat different perspective, Thompson and Thomas concluded that although such suits generate significant litigation-related agency costs, they also produce large enough reductions in managerial agency costs to "deserve a seat at the table of corporate governance."30
To reach our own conclusion as to whether the costs associated with merger-related class actions outweighed their benefits, we undertook to examine intensively all merger-related class actions filed in Delaware Chancery Court with respect to mergers announced between January 1, 1999, and December 31, 2001.31 We chose to cover a three-year period, rather than the two-year period of the Thompson and Thomas study, to increase the statistical reliability of our results. At the same time, we thought that it would be useful for the period of our study to overlap with that of Thompson and Thomas's study. We also anticipated (correctly, as it turned out) that this was the latest three-year period in which almost all merger-related class actions that were filed had been resolved. Finally, the relevant principles of Delaware law and corporate practices during this period remained relatively stable. We would not need to adjust our observations to take account of possible changes in Delaware law or corporate practices.
We based our study on the neo-classical assumption that plaintiffs' attorneys are rational economic actors, who because they generally operate without meaningful client control, make litigation-related decisions primarily to advance their own economic interests.32 However, unlike the Delaware courts, which additionally appear to believe that they can curb attorney opportunism through judicial monitoring of settlements and fee awards, our model assumed that other aspects of Delaware law largely shaped plaintiffs' attorneys' litigation-related decisions. More specifically, we assumed that an analysis of the pleading standards of Delaware merger law, including burden of proof and standard of review, would allow us to predict the kinds of mergers most likely to lead plaintiffs' attorneys to file class-action complaints, which merger-related lawsuits they were most likely to prosecute actively, and when and on what terms those lawsuits were most likely to be settled.33
We then examined all mergers within our research universe and all class actions filed with respect to those mergers to see if they bore out our predictions. We made no predictions as to the effectiveness of judicial monitoring, but we did attempt to assess how effective it had been. Our overall conclusions are that Delaware law relating to mergers and class actions created a litigation environment that was rife with potential for opportunistic behavior by the plaintiffs' bar; that plaintiffs' attorneys generally responded by behaving opportunistically; and that Delaware's courts did not effectively protect corporations or their shareholders from the resulting litigation-related agency costs.34
The balance of this paper proceeds as follows. Part II describes the different categories of mergers that we studied, the principles of Delaware law applicable to those mergers and to merger-related class actions, and the practices that Delaware corporations generally follow in arranging such mergers. It also contains our predictions as to what mergers are most likely to be challenged by class actions, which class actions are most likely to be prosecuted, and on what basis those class actions are most likely to be resolved. Part III describes our research methodology and describes and assesses our findings. Part IV concludes.
II. DELAWARE LAW AND LITIGATION INCENTIVES
Delaware law can be viewed as initially dividing mergers, acquisitions, and other business combinations-which we shall refer to collectively as "mergers"-into two basic categories and then subdividing mergers in each of those categories into subcategories. The basic division is between mergers negotiated at arm's length and those that involve a conflict of interest. Such conflicts exist either because an officer, director, or controlling shareholder of the company being acquired is, or is affiliated with, the other party to the merger or because the merger allegedly involves some side deal with an officer, director, or controlling shareholder of the target company. Within the arm's-length category, Delaware courts apply different standards of review to mergers that involve a sale of control, mergers directed at fending off hostile takeover bids, and mergers that involve neither a sale of control nor any defensive action. Within the conflict-of-interest category, Delaware courts apply different standards of review to mergers in which the acquirer is a controlling shareholder ("squeeze outs"), mergers in which officers or directors of the target company have an ownership stake in the acquirer but do not control the target ("management buy outs" or "MBOs"), and mergers in connection with which an officer or director of the company being acquired has some other financial interest that conflicts with the interests of that company's public shareholders. Analysis is further complicated by variations in standards of review in Delaware courts, which depend on whether the merger involves only a tender offer or whether it also involves a merger consummated pursuant to section 251 of the Delaware General Corporation Law ("DGCL") (a "statutory merger"), a merger consummated pursuant to section 253 of the DGCL (a "short form merger"), or some other action by the target company's board of directors.
A. Mergers Negotiated at Arm's Length
A merger negotiated at arm's length with an unrelated acquirer is unlikely to involve financial unfairness. Neither the officers nor the directors of the target company have any incentive to agree to a transaction that is not in shareholders' best interests. Both the requirement for approval from a majority of the target's shareholders and the fact that an unrelated shareholder's "friendly" tender offer will succeed only if a majority of the target's shareholders tender their stock further reduce the potential for abuse of shareholders' interests. Thus, one might anticipate that all board decisions to approve mergers involving unrelated acquirers will be protected from judicial scrutiny by the business judgment rule-that is, that a court will presume that the board agreed to the merger in good faith and after reasonable investigation. A class-action complaint challenging such a merger then would be unlikely to survive a motion to dismiss unless the complaint alleged with particularity facts that, if true, would raise a reasonable doubt as to whether the presumptions of the business judgment rule applied.35 A complaint asserting no more than generalized claims that the board's decision was unwise, imprudent, or uninformed clearly would not survive.
Delaware courts, however, do not treat all board decisions to approve arm's-length mergers as protected business judgments. Rather, they distinguish between mergers that involve sales of control and those that do not. Unless plaintiff can show that the merger or some provision of the merger agreement serves a defensive purpose, the business judgment rule will insulate from judicial scrutiny a decision to approve a merger that does not involve a sale of control.36 But if a merger involves a sale of control, Delaware courts will subject the board's decision to "enhanced scrutiny."37 More specifically, they will require the defendant directors to bear the burden of proving that they acted reasonably to seek the transaction offering the best value reasonably available to the target company's shareholders.38 The business judgment rule will protect the directors' approval of the sale only if they meet that burden.
Whether a merger involves a sale of control thus is a matter of critical doctrinal importance. The Delaware courts' rationale for holding that sales of control are subject to enhanced scrutiny is that shareholders' voting rights are of overriding importance because they constitute the principal mechanism of shareholder participation in corporate governance. Consequently, shareholders should be compensated if and when the corporation enters into a transaction -a sale of control-that will result in the loss of these governance rights.39 Any merger in which shareholders receive only cash or debt instruments in exchange for their stock thus qualifies as a sale of control,40 as does any merger in which the target's shareholders receive stock in a corporation that, after the merger, will be controlled by a single person, entity, or cohesive group.41 On the other hand, when a corporation with no controlling shareholder participates in a merger in which its shareholders will receive stock for their stock and control of the surviving corporation will continue to reside in "a large, fluid changing and changeable [public] market," no sale of control will be deemed to have occurred. The business judgment rule then will apply42 unless the merger serves a defensive purpose.
Delaware courts review stock-for-stock mergers that serve a defensive purpose under what is known as the Unocal standard.43 They require the directors of the corporation taking the defensive action to demonstrate both that "a danger to corporate policy and effectiveness existed" and that the defensive measure is "reasonable in relation to the threat posed."44 However, if "a defensive measure is not draconian . . . because it is not either coercive or preclusive," then in most circumstances a Delaware court will find it to be reasonable.45
In relation to arm's-length mergers that involved sales of control, our hypothesis was that the applicable standard of review and burden of proof would make it attractive for plaintiffs' attorneys to file class actions alleging that the directors of the selling corporations had breached their Revlon duties. A complaint that alleged the facts of the proposed sale of control and made such a generalized allegation would shift to the defendant directors the burden of proving that they had met their Revlon duties. That alone, we estimated, would provide plaintiffs' attorneys with sufficient leverage in some cases to extract a nonmonetary settlement. However, because nothing would prevent both the target company's board and its shareholders from seeking the highest value reasonably available, and because Revlon duties are in a sense self-enforcing,46 we also hypothesized that few class actions challenging arm's-length sales of control would result in any monetary recovery for the plaintiff class and that, in at least some cases, defendants convinced that they had not breached their Revlon duties would signal their intent to contest plaintiffs' claims. Moreover, the biggest potential payoff for a plaintiffs' attorney filing a class action challenging a sale of control was likely to occur where, for one reason or another, a competing bidder emerged and the company thereafter was sold for a higher price. In that event, a plaintiffs' attorney-even if her original claim that Revlon duties had been breached had no reasonable basis-nonetheless would be in a position to claim a share of the credit for the higher price and a fee award for her efforts.
Consequently, we hypothesized, Delaware law made it moderately attractive for plaintiffs' attorneys to file suits challenging sales of control. They could do so at relatively low cost; they might thereafter find it easy to extract nonmonetary settlements (and related fee awards) even if no wrongdoing could be found;47 and they might even "hit the jackpot" if a competing bidder emerged. At the same time, especially because defendants who were confident that they had met their Revlon duties often would be in a position to resist successfully such claims without incurring substantial litigation costs,48 we anticipated that plaintiffs' attorneys would elect not to prosecute actively a significant proportion of the complaints that they filed challenging sales of control. Rather, unless a competing bidder emerged (or some other unanticipated event or revelation occurred), we expected that plaintiffs' attorneys would dismiss voluntarily a significant proportion of the complaints challenging such mergers that they filed.
We also recognized that it occasionally is difficult to determine whether an arm's-length merger involves a sale of control. Examples include mergers in which the target's shareholders receive some combination of stock and cash (or debt) and mergers in which it is unclear whether a single person, entity, or cohesive group will control the surviving corporation.49 When a class action is filed challenging such a "hybrid" merger, some initial skirmishing is likely over whether Revlon duties apply. This will make challenging a hybrid merger less attractive to plaintiffs' attorneys than challenging a clear sale of control but more attractive than challenging a stock-for-stock merger that does not involve a sale of control. On the other hand, the board that negotiates a hybrid merger may well be more concerned with factors such as "strategic fit" rather than with obtaining the highest value reasonably available, especially if it does not believe that Revlon duties apply. Thus, a monetary recovery may be more likely if a class action is filed challenging a hybrid merger and plaintiffs prevail on the sale of control issue. On balance, that possibility led us to predict that suits challenging hybrid mergers were somewhat less likely to be filed than suits challenging clear sales of control, but considerably more likely to be filed than suits challenging stock-for-stock mergers. We also predicted that successful, Revlon-based challenges to hybrid mergers were more likely than challenges to sales of control to result in monetary recoveries.50
The transactional form of an arm's-length merger is not likely to affect the foregoing analysis. Whether a sale of control involves a merger or a tender offer is unlikely to be significant because, so long as the target company's board takes some action to facilitate the transaction, the action it took will be subject to enhanced scrutiny under Revlon. Similarly, if a board agrees to a stock-for-stock merger that does not involve a sale of control, its actions will be reviewed under the business judgment rule in all but one set of circumstances-when a court could conclude that the target's board agreed to the merger or to deal protection measures primarily to protect the target company from an uninvited takeover bid by some other aspiring acquirer.
As noted above, Delaware law allows boards of directors considerable freedom to take defensive actions, so long as they are not preclusive or coercive.51 Although challenges to defensive measures usually fail,52 plaintiffs' attorneys still may find it attractive, for entirely pragmatic reasons, to file complaints challenging stock-for-stock mergers that serve defensive purposes. Even if it has installed impregnable takeover defenses,53 the board of a target company will often give in to the pressures of the marketplace and abandon its defenses when a hostile bid gets high enough. A plaintiffs' attorney who has filed a class action challenging that companys' takeover defenses then can claim credit for the target's surrender and demand a fee equal to at least a modest percentage of the higher price that shareholders receive, which may, in total, amount to hundreds of millions or even billions of dollars. Thus, we predicted that plaintiffs' attorneys would find it attractive to file class actions challenging stock-for-stock mergers when, at the time the merger was announced or shortly thereafter, a competitive, hostile bid for the target corporation had been made or seemed reasonably likely. If a contest for control then developed, the plaintiffs' attorney could continue to prosecute that claim but would, in all probability, largely free-ride on the litigation efforts of the hostile bidder. Such a suit, however, was likely to lead to the plaintiff class's realizing a very substantial monetary benefit and to plaintiffs' attorneys' receiving a very substantial fee award. Where no contest for control developed, though, we anticipated that plaintiffs' attorneys would elect to dismiss voluntarily complaints challenging defensive mergers, since they would find it difficult to succeed on the merits.
We also anticipated that a similar dynamic would govern challenges to stock-for-stock mergers that did not involve either a sale of control or any defensive measure. Plaintiffs' attorneys might find it attractive to file class actions challenging such mergers, especially where the company being acquired was large in size, in the hope that a contest for control would develop after the merger was announced and the attorneys filing such suits then would realize windfall profits largely as a consequence of the efforts of the competing bidder. On the other hand, we anticipated that absent the emergence of a competing bid, plaintiffs' attorneys would voluntarily dismiss virtually all such suits, since a complaint challenging such a merger would be vulnerable to a motion to dismiss based on the business judgment rule and thus would provide plaintiffs' attorneys with very little litigation leverage.
B. Mergers Involving Conflicts of Interest
As we noted above, Delaware courts apply different standards of review to squeeze outs, MBOs, and mergers involving other conflicts of interest than they do to mergers negotiated at arm's length. As concerns squeeze outs, Weinberger v. UOP, Inc.54 holds that a controlling shareholder bears the burden of proving that the merger is "entirely fair."55 This is commonly described as the most rigorous standard of review applied under Delaware corporate law.56
Weinberger explains that entire fairness includes two elements, fair dealing and fair price, but provides no real guidance as to how those two elements interact.57 However, the court's opinion hints strongly that a controlling shareholder's best approach is to cause or allow the target company's board of directors to appoint an SNC with the authority to negotiate the terms of the squeeze out on behalf of the target's public shareholders.58
Not surprisingly, it soon became standard operating procedure for the board of a controlled corporation to create an SNC whenever a controlling shareholder announced its intent to squeeze out the public shareholders and for the SNC then to retain legal and financial advisors to assist it in negotiating with the controlling shareholder.59 Practitioners generally viewed such actions as necessary, though not sufficient, to demonstrate that a squeeze out was entirely fair.
Kahn v. Lynch Communications Systems, Inc.60 largely resolved a related uncertainty-how proof that an SNC had effectively represented the interests of public shareholders would affect the standard of review and burden of proof that Weinberger had established. Lynch I held that proof of fair dealing, meaning that the terms of a squeeze out had been negotiated at arm's length by an SNC with real bargaining power, would shift to a shareholder challenging the merger the burden of proving that any price to which the SNC had agreed was not entirely fair.61
However, if a controlling shareholder decided to attempt to effect a squeeze out using a tender offer without first obtaining the support of the controlled subsidiary's board of directors, then Weinberger would not apply. Rather, Solomon v. Pathe Communications62 held that a controlling shareholder had no duty to offer a fair price; its fiduciary obligations were limited to avoiding "coercion or disclosure violations."63 However, the court did not explain when it would deem a controlling shareholder's tender offer to be coercive, nor did any other reported decision discuss that issue until late in 2001. Thus, as to this one category of mergers-a squeeze out effected by a tender offer for which the support of the subsidiary's board of directors had not been obtained-the governing legal principles remained somewhat uncertain during most of the period under review.
On the other hand, for most of that same period, it seemed clear that if a controlling shareholder decided to effect a squeeze out by means of a short-form merger (whether as the sole step in the squeeze out, if it owned more than 90 percent of the target, or as a second step following a first-step tender offer), then the controlling shareholder had a fiduciary obligation to prove that the short-form merger was entirely fair. Only late in 2001, in Glassman v. Unocal Exploration,64 did the Delaware Supreme Court reverse itself and hold that a controlling shareholder had no obligation to prove that a short-form merger was entirely fair.65 Consequently, as concerned almost all the squeeze outs included in our study, the operative planning assumption was that, at some stage in the process, entire fairness would be the relevant standard of review.66
As Professor Rock has pointed out, Weinberger and other Delaware cases decided prior to 1990 also provided "substantial guidance on how to structure a management buyout transaction."67 As when a squeeze out is proposed, the board of the target company should appoint an SNC composed of independent directors, which should then retain its own investment bankers and legal counsel and enter into negotiations with the MBO group.
However, at this point in the process, the responsibilities of an SNC will differ somewhat from those of an SNC appointed to negotiate a squeeze out. Almost every MBO also involves a sale of control.68 Thus, an SNC charged with negotiating an MBO must keep in mind its duties under Revlon. This both complicates and simplifies the SNC's task. On one hand, the SNC cannot limit its efforts to negotiating with the management group. It also must test the market, in some reasonable fashion, to determine whether the management group's offer represents the highest value reasonably available.69 Moreover, if a third party expresses interest in acquiring the target, the SNC has a duty to provide it with information comparable to that available to the MBO group and otherwise to deal with that party in an evenhanded fashion.70
On the other hand, an SNC that meets its obligations under Revlon generally will find it easier than an SNC that negotiates a squeeze out to prove that the price it negotiated was fair.71 In the setting of an MBO, so long as the SNC proceeded reasonably, the market test itself will constitute persuasive evidence that the target's shareholders are receiving the highest value reasonably available-a price that, by definition, also will be deemed to be fair.
In the case of a squeeze out, however, comparable, market-based evidence will almost never be available because, under Delaware law, a controlling shareholder has no obligation to sell her stock.72 Moreover, a controlling shareholder that is proposing a squeeze out almost always will make clear that she has no desire to sell. Consequently, squeeze outs involve greater litigation risks for defendants. No matter how effectively an SNC has done its job, the question of whether the price it negotiated is fair ultimately must be resolved on the basis of the (inevitably competing) opinions of plaintiffs' and defendants' experts.73 That is a battle that defendants sometimes will lose.74
Despite these differences, Delaware law promotes a similar transactional dynamic with respect to both squeeze outs and MBOs. In both situations, it effectively discourages the potential acquirer from initially placing its best offer on the table. Both a controlling shareholder and an MBO group will anticipate that the target company's board will appoint an SNC. Both will also know that, whether the SNC actually bargains vigorously on behalf of the public shareholders or only goes through the motions, evidence that the SNC succeeded in negotiating a higher price will make it much more likely that a court will find that the SNC met its fiduciary obligations and that the price to which it agreed was entirely fair. Consequently, the price agreed to by SNCs in most (if not all) squeeze outs and MBOs will be higher than the price initially offered, regardless of whether any given SNC was an effective advocate for the interests of the target's public shareholders.75
Plaintiffs' attorneys, we hypothesized, would be familiar with this transactional dynamic and would realize that it provides them with strong incentives to file class actions challenging both squeeze outs and MBOs as soon as they are announced. A "boilerplate" complaint that simply recites the structure of the proposed transaction, asserts that it involves a conflict of interests, and alleges that the price offered is unfair or inadequate will almost always survive a motion to dismiss and will also shift to defendants the burden of proving that the proposed merger is entirely fair, in the case of a squeeze out, or that the members of the SNC were independent and that they obtained the highest price reasonably available, in the case of an MBO.
Moreover, plaintiffs' attorneys' bargaining power would derive from far more than their ability to impose litigation costs on defendants. In the case of almost every squeeze out and MBO, a plaintiffs' attorney could anticipate that the price finally agreed to by the SNC would be higher than the price originally offered by the controlling shareholder or MBO group. Consequently, by supplementing her complaint with a presentation to the SNC of her reasons for believing that the target company's stock is worth more than the price originally offered, a plaintiffs' attorney also could put herself in a position to claim some credit for whatever increase in price was then negotiated by the SNC and to demand compensation for her efforts.
Of course, neither a class-action plaintiff nor her attorney has a right to present information to an SNC, since the committee, acting on behalf of the board, is charged with the management of the corporation.76 Nonetheless, we found that SNCs regularly invite plaintiffs' attorneys to present their arguments regarding valuation before those committees finally decide on what terms they will approve a squeeze out or an MBO.77
SNCs may follow this practice because it can yield significant tactical benefits. As described in the paragraphs that follow, Delaware law gives plaintiffs' attorneys strong incentives to sign a Memorandum of Understanding ("MOU") recording their acquiescence in whatever deal an SNC has negotiated, so long as (a) that deal involves a price greater than the price originally offered, and (b) defendants acknowledge in the MOU that plaintiffs' attorneys efforts "contributed" to the price increase. With such a MOU in hand, an SNC can be highly confident that any deal to which it has agreed is unlikely to be subject to further challenges. On the other hand, if the SNC did not provide plaintiffs' attorneys with an opportunity to appear, those attorneys would have no incentive to acquiesce in the deal negotiated by the SNC, since by doing so they would relinquish their claims without having established a clear basis for claiming entitlement to a fee award. Moreover, especially in a case in which a plaintiffs' attorney reasonably could argue that the improved price negotiated by an SNC still was unfair, that attorney would have a strong incentive to continue to challenge the merger. Thus, inviting a plaintiffs' attorney to "participate" in its deliberations may allow an SNC to purchase what might be termed "litigation insurance" at relatively modest cost.78 At least from the shareholders' perspective, though, so proceeding poses a real danger. It could be a prelude to an implicitly collusive settlement in which plaintiffs' attorneys, in exchange for defendants' virtual guaranty of a fee award, agree to sign off on merger terms that at least arguably are unfair and that they might otherwise be successful in challenging.
Moreover, where a plaintiffs' attorney indicates that she is prepared to acquiesce in an SNC's pricing decision, defendants typically are prepared to acknowledge in the MOU recording her acquiescence that the efforts of plaintiffs' attorneys contributed to some degree to whatever the improvement in price the SNC has negotiated.79 This virtually assures plaintiffs' attorneys of a generous fee award.80 Thus, by agreeing to settle for the price negotiated by an SNC, a plaintiffs' attorney can obtain an almost-certain, substantial benefit and will incur little in the way of costs.81
At the same time, the structure of Delaware law operates to discourage plaintiffs' attorneys from challenging the fairness of any revised transaction to which an SNC has agreed, whether or not those attorneys believe an SNC has effectively protected shareholders' interests. A plaintiffs' attorney who insists on challenging a price agreed to by an SNC inevitably will incur substantial costs without any assurance that she will realize any benefits from her efforts. Unless the plaintiffs' attorney can show that the members of the SNC were not independent, she will bear the burden of proving that the improved price to which the SNC agreed nonetheless is unfair.82 In virtually every case, the SNC's decision will be supported by a fairness opinion from some prominent investment bank. Thus, to prove unfairness, the plaintiffs' attorney will face the added burden of convincing the court that the investment bank was wrong. At a minimum, a plaintiffs' attorney will have to incur substantial litigation costs to mount such an effort. Moreover, since no MOU will have been signed and defendants therefore will not have acknowledged any positive contribution by plaintiffs' attorneys, if those attorneys cannot convince the court that the improved price is unfair, they are likely to receive no compensation whatsoever for their litigation efforts.
As concerns both squeeze outs and MBOs, then, Delaware law provides plaintiffs' attorneys with strong incentives to file class-action complaints and to try to advance claims of unfairness up to the point at which an SNC agrees to an improved price.83 Thereafter, Delaware law discourages plaintiffs' attorneys from continuing to challenge any improved deal to which an SNC has agreed, as the probable costs of continuing to litigate will, in almost all cases, far outweigh the probable benefits.84 Consequently, we predicted, plaintiffs' attorneys would agree to settle virtually all class actions challenging squeeze outs and MBOs in connection with which an improved price was negotiated by an SNC after suit was filed. We also recognized that if we found either that plaintiffs' attorneys had chosen to challenge a substantial portion of the improved deals negotiated by SNCs or that plaintiffs' attorneys were demanding fee awards equal to a substantial percentage of the price increases to which SNCs had agreed,85 that would strongly suggest that plaintiffs' attorneys' litigation decisions were motivated largely by a sense of loyalty to the plaintiff class, rather than by concern for their own financial interests.
We found it considerably more difficult to predict how litigation would proceed where class actions were filed challenging mergers involving other alleged conflicts of interest. Often, whether an actionable conflict exists will not be clear. Thus, a plaintiffs' attorney is likely to find it difficult to predict at the outset whether her claim will be governed by the duty of loyalty or the business judgment rule. This uncertainty seemed likely to make merger-related class actions alleging other conflicts significantly less common than actions challenging squeeze outs and MBOs, where both governing legal principles and customary corporate practices were well established. It also seemed that the presence of such uncertainty made it significantly more likely that where such other conflicts of interest were alleged, litigation would involve a judicial decision addressing the merits of whatever claim plaintiffs' attorneys had made.
III. RESEARCH METHODOLOGY AND FINDINGS
A. Introduction and Hypotheses
The discussion in Parts I and II yields two broad alternative hypotheses as to what drives shareholder litigation in Delaware. Both start with the neo-classical economic assumption that plaintiffs' attorneys are rational economic actors who act primarily with a view to advancing their own economic interests. The first hypothesis is that the structure of Delaware law channels the self-interest of the attorneys so that they are genuine advocates of shareholders' best interests; that is, these suits are brought by attorneys who are acting primarily in the interests of shareholders and who thereby serve as an effective "police force" to ensure compliance with corporate fiduciary duties. This appears to be the vision of the Delaware courts. The courts acknowledge that plaintiffs' attorneys may instead act in ways that primarily promote their own interests rather than the interests of their nominal clients. But, the courts claim, they are prepared to monitor the activities of plaintiffs' attorneys, primarily through reviews of settlements and of fee awards, so as to keep those attorneys focused on their beneficial role and discourage them from behaving in an opportunistic manner. We label this the shareholder champion hypothesis.
We offer a second hypothesis that, we believe, applies to merger-related class actions: Delaware substantive law relating to pleading standards, burdens of proof, and standards of review, when overlaid on the system that Delaware courts have created to encourage plaintiffs' attorneys to initiate and maintain private litigation, provides plaintiffs' attorneys with strong incentives to act largely in their own interests rather than in the best interests of shareholders, and plaintiffs' attorneys have responded to these incentives in a predictably self-interested fashion. Implicit in this hypothesis is the view that courts' restraint on attorneys' behavior, through review of settlements and fee awards, has been scanty and ineffective in curbing such self-interested behavior. We label this the self-interested litigator hypothesis.
To some extent, the two hypotheses have similar implications. Both the beneficial and the not-so-beneficial forms of litigation will be drawn to mergers where the pleading grounds are stronger, where the transactions are larger (either because of greater potential injuries incurred by shareholders, or because of the greater willingness of defendants to pay "modest" fees to settle nuisance suits), and where the initial premium offered to target company shareholders is smaller (again, either because shareholders may have suffered greater harm, or the plaintiffs' attorneys may simply hope for more sympathy from a judge and thus a greater willingness of defendants to settle). But, given the relatively low costs of initial filings, the self-interested litigator hypothesis would predict faster and more frequent filings when cases are filed, as well as opportunistic efforts to free ride on the efforts of others (such as SNCs and competing bidders) when possible. Further, the self-interested litigator hypothesis would predict ready acquiescence in any improvements that SNCs achieve, regardless of whether any given SNC was diligent in promoting shareholders' interests, and few post-improvement challenges.
To test these hypotheses, we initially collected data for the years 1990-2001 with respect to mergers where the target company was publicly traded and incorporated in Delaware and the deal value was in excess of $100 million.86 Next, we collected information on those stockholder class-action lawsuits that were filed against a target company following the announcement of a merger. The information collected includes the date of the filing of the suit (or of the first complaint if multiple complaints were filed), the allegations in the complaint, the legal outcome, the terms of the outcome (such as the terms of any settlement), the fees (if any) awarded to the plaintiff law firm, and other relevant events. Because of the substantial data collection burden involved in identifying, matching, and describing the legal information where suits have been filed, we have restricted our analyses to the years 1999-2001.
We initially examined the characteristics of those mergers (104) that were challenged by lawsuits and those (460) that remained unchallenged. These analyses are discussed in Section B. They indicate that the bringing of shareholder class-action lawsuits in Delaware is a systematic process-it is clearly not random-and that it follows the patterns suggested above. The results of these suit-initiation analyses, however, do not allow us to distinguish between our two broad hypotheses.
We then examined the patterns of the 104 suits that were actually brought. In Section C we describe broad patterns. Especially worth noting among these are the high percentage of cases in which an initial complaint was brought within two business days and the pattern of multiple filings of suits. In line with Thompson and Thomas, we believe that multiple filings, as well as early filings of complaints, suggest opportunistic behavior by the plaintiffs' bar. In Section D we summarize our detailed examination of the pattern of settlements and dismissals against the background of the categories of merger. That pattern, we believe, is broadly supportive of the self-interested litigator hypothesis. In Section E we discuss the five (of the 104) cases where "real" clients, with substantial financial interests, were represented by attorneys from outside the "traditional" Delaware plaintiffs' bar. The differences in the ways that these five cases were pursued, as compared with the pattern described by the remaining (99) cases that were litigated by members of the "traditional" plaintiffs' bar, are again supportive of the self-interested litigator hypothesis. In Section F we evaluate how effectively the Delaware courts have monitored settlements and fee awards in the cases brought by "traditional" plaintiffs' attorneys and find that the courts have not been effective monitors. Finally, in Section G we provide a summary and evaluation of these empirical results and their implications.
B. The Influences on Legal Challenges to Mergers
The Thomson database yielded 564 mergers for the years 1999-2001 where the target company was a publicly traded Delaware company and where the value of the transaction exceeded $100 million.87 Of these 564 mergers, 104 were challenged. We initially asked whether the pattern of challenges was systematic-whether there were features of these mergers that tended to increase the likelihood that a merger would be challenged. Equivalently, these features would be influences on the incentives of plaintiffs' lawyers to initiate and maintain shareholder class-action lawsuits.
Following the discussion above, we focused on four potential features of mergers that might influence the likelihood that the merger would be challenged: First, all-cash mergers, as sale-of-control transactions, would be more likely to attract stockholder class-action lawsuits than would mergers where the acquirer offers stock to the target company's shareholders. Second, lawsuits should be more likely where the transactions are larger88 (which could indicate greater potential injury to shareholders and, in any event, would increase the likely size of recoveries/settlements89). Third, suits should be more likely where the share price offered by the acquiror (as compared to the pre-offer market price) represents a relatively small premium (and thus, again, shareholders may face greater injury, and plaintiffs' attorneys may find a more sympathetic judicial ear). Fourth, the findings of Thompson and Thomas indicate that self-dealing abuses are also likely to trigger lawsuits.90 Two likely indicators of self-dealing would be (a) whether the acquiror owned a stake in the target company (and thus might influence the target's senior management and board of directors to accept an unduly low offer price); and (b) whether the senior management (and/or board) have come to a prior understanding with the acquiror as to post-acquisition positions and/or remuneration (and, thus, might be unduly prone to accept this acquirer's offer and reluctant to seek better terms for public shareholders-perhaps by "shopping" the company around to other potential acquirers-for fear of souring a deal with this acquiror).91
As our discussion above92 indicates, we recognize that these variables have dual interpretations: they could support either the shareholder champion hypothesis or the self-interested litigator hypothesis. Consequently, they will not allow us to differentiate between the two. Nevertheless, they give us a first look at whether the overall pattern of litigation is purposive and systematic or whether it is largely random and inexplicable.
1. Differences in means and medians
The available data for 1999-2001 are arrayed in Table 1, according to whether a merger announcement was shortly followed by one or more class-action lawsuits or whether the transaction was unchallenged; the data for the individual years are arrayed in Tables 2-4. In addition, the means of the variables-for the mergers that attracted lawsuits, and the mergers that did not-are portrayed graphically in Figures 1-8.
We begin by describing the overall sample. As can be seen in Table 1, there were 564 qualifying93 mergers over the years 1999-2001, of which 104 (18.4 percent) were followed by at least one stockholder class-action lawsuit and 460 (81.6 percent) were unchallenged. The mergers that attracted lawsuits tended to be larger, as predicted.94 This is true for measurements involving the mean as well as the median, and for size measured by value of the deal,95 the target's annual sales, the target's net income, the target's total assets, and the target's value of common equity.96 The offer-price premium97 was smaller for the mergers that attracted lawsuits than for the mergers that did not, as predicted.
The mergers that attracted lawsuits were more likely to be all-cash deals, as predicted. Almost two-thirds of the lawsuits (67 out of 104, or 64.4 percent) followed all-cash deals, whereas less than a third (131 of 460, or 28.5 percent) of the lawsuit-free deals were all cash.98
Stockholder class-action lawsuits also tended to occur in mergers where the acquiror had a prior ownership stake in the target company or there were appearances of self-dealing, as predicted.99 Almost two-thirds of the lawsuits (68 out of 104, or 65.4 percent) followed mergers where the acquiror already had an ownership stake in the target company or there was self-dealing, whereas less than 5 percent (21 out of 460, or 4.6 percent) of the lawsuit-free mergers involved such circumstances.
When the means and medians for the individual years 1999, 2000, and 2001 are examined, similar patterns emerge with respect to all of these variables, although the levels of statistical significance are more varied.
2. Regression analysis
Though differences in means and medians provide a useful "first cut" at the data, these univariate comparisons do not allow for covariation and multivariate analysis. Accordingly, we now turn to multivariate (logit) regression analysis.100
In Table 5 we show the results of logit regressions for the combined years 1999-2001, and in Tables 6-8 we show the results for each year separately. The dependent variable is a 1,0 lawsuit variable (a merger where one or more lawsuits were filed = 1; a merger where no lawsuit was filed = 0), and a variety of the characteristics of the mergers are the independent (explanatory) variables.
As can be seen in Table 5, for the combined years, the most powerful explanatory variable is whether the acquirer had a prior ownership stake in the target company (or there were other appearances of prior self-dealing). A variable that is related to the size of the deal or the size of the target company (whether in natural numbers or in logarithms) is also generally (but not always) significant.101 Further, an all-cash deal tended to attract lawsuits, as did the payment of a lower premium paid for the target's shares; both variables are consistently significant.
The results for the individual years, shown in Tables 6-8, reinforce the impression that an acquiror's prior stake and the size of the transaction tended to be the powerful influences on whether a lawsuit was initiated. The other variables' influences are more spotty (in terms of statistical significance), although the plus and minus signs of the coefficients (and thus the predicted direction of effects in terms of encouraging or discouraging lawsuits) are consistent.
3. A summing up
The data on means/medians and the logit regression analyses yield consistent inferences: a prior ownership stake or other appearances of conflicts, a larger size of the transaction or of the target, the presence of an all-cash transaction, and a lower initial premium offered to the target's shareholders all influenced significantly the likelihood that a merger would be challenged with one or more shareholder class-action lawsuits. Although these results do not allow us to distinguish between our two broad hypotheses, they do indicate that the initiation of these lawsuits was systematic in sensible ways and was far from random.
C. A Brief Summary of the Lawsuits
As described above, there were 104 mergers that attracted at least one class-action lawsuit. Of these suits, 48 settled, 54 were dismissed, and 2 are still pending as of this writing. Of the settlements, 31 involved claimed monetary recoveries for the plaintiffs, and 17 involved only nonmonetary terms. These outcomes are described graphically in Figure 9. We will first describe some of the characteristics of the overall sample of the 104 cases and then of the 48 settlements and 54 dismissals.
1. Speed of filing
An immediately striking fact is the speed with which these complaints were filed: Of the 104 challenged mergers, 77 (74.0 percent) attracted their initial lawsuit within one business day of the merger announcement. An additional 7 mergers (6.7 percent) attracted their initial lawsuit on the second business day after the merger announcement. Thus, for the mergers that did attract lawsuits, almost three-quarters of the time the initial complaint was filed within one business day of the merger announcement, and over 80 percent of the time it was filed within two business days.102
As do Thompson and Thomas,103 we believe that such rapid/early filings of complaints suggest opportunistic behavior by the plaintiffs' bar.
2. Numbers of complaints filed
For these 104 challenged mergers, multiple suits filed by a law firm that is a member of the "traditional" Delaware plaintiffs' bar were the norm. The mean number of suits per challenged merger was 5.7, the median was 5, and only 12 challenged mergers (11.7 percent) were the subject of only a single complaint.
Again following Thompson and Thomas, we believe that multiple complaints suggest opportunistic behavior by the plaintiffs' bar.
3. Settled and dismissed cases
The outcomes of the lawsuits were close to evenly split between settlements (48, or 47.1 percent of the 102 cases with known outcomes104) and dismissals (54, or 52.9 percent of 102). None of the cases was litigated to a judgment on the merits in favor of the plaintiffs. However, a second striking fact emerges from a detailed examination of the dismissals: Of the 54 dismissed cases, 51 (94.4 percent) of the dismissals were dropped before there was a judicial decision that addressed the merits of the lawsuit; only three of the dismissals followed an adverse judicial decision addressing the merits. Moreover, very few of the 51 cases that plaintiffs' attorneys dismissed voluntarily involved significant litigation efforts by plaintiffs following the filing of a complaint or, in most instances, multiple complaints. This further suggests to us that plaintiffs' attorneys' decisions concerning when to file and whether a case warranted active prosecution probably were governed more by those attorneys' economic self-interest than by client-oriented processes and thus provides support for the self-interested litigator hypothesis.
For the 48 challenged mergers where there were settlements, in 43 instances (89.6 percent) the first complaint was filed within two business days of the merger announcement. In these cases, the mean number of complaints filed was 7.3, and there were only 3 cases (6.2 percent) in which only a single complaint was filed.
For the 54 challenged mergers where all of the suits were eventually dismissed, in 41 instances (75.9 percent) the first complaint was filed within two business days of the announcement. In these cases, the mean number of complaints filed was 4.4, and there were 9 cases (16.7 percent) in which only a single complaint was filed.
4. Monetary recoveries
Of the 48 settlements, in almost two-thirds (31, or 64.6 percent), the plaintiffs' attorneys claimed to have achieved or contributed to monetary recoveries for the class, in terms of improvements in the payments that the target's shareholders had received or would receive. These claimed recoveries averaged $163 million, but ranged from a low of $3 million to a high of $1,250 million, with a median of $50 million. The size of the claimed recovery was positively and significantly related to the announced value of the transaction and to the various size measures of the target; but after the value of the transaction was controlled for, no other variable played a significant role in explaining the size of the recovery. When expressed as the percentage of improvement in the price received by the target shareholders, these monetary recoveries yielded a 15.6 percent price improvement, with a range of 0.8 percent to 50.0 percent and a median of 12.6 percent. However, consistent with the self-interested litigator hypothesis, we found that plaintiffs' attorneys frequently were able to free ride on the improved terms negotiated by SNCs or on the price improvements that resulted from competing bids, that they rarely claimed a major share of the credit for the improvements, and that they never persisted in challenging the terms negotiated by an SNC or the terms proposed by a competing bidder.105
5. Legal fees
For the 48 cases that settled, information on the legal fees received by plaintiffs' attorneys was generally available. We portray this information in Figure 9. As can be seen, the aggregate fees received, as well as the hourly rates, were far higher for the settlements where monetary recoveries were claimed than for the settlements where only nonmonetary terms were achieved. When viewed on an hourly-rate basis, the fees also seem rather high, especially when one takes into account the fact that we computed hourly rates on the basis of every lawyer-hour worked, without distinguishing between senior partners and junior associates, and in some cases also included hours worked by paralegals. For settlements that involved no monetary recovery, legal fees averaged $492 an hour, and the median fee award was equal to $472 an hour. For settlements that involved a monetary recovery, the average fee awarded was equal to $1,800 per hour for every attorney-hour worked and the median was $1,240 per hour. Moreover, our examination of all merger-related class actions filed in 1999-2001 suggests that the attorneys who brought these cases did not face much in the way of contingency risk.106 Thus, our results suggest that litigating merger-related class actions in Delaware Chancery Court appears to be a lucrative area of practice for the plaintiffs' bar.
We attempted to explain the pattern of legal fees received, where, alternatively, the aggregate fee per case and the hourly rate were the dependent variables. Preliminary analysis indicated that, for those settlements where there were claimed monetary recoveries, the aggregate fee and the hourly rate were only mildly (and insignificantly) positively related to the claimed size of the monetary recovery.107 We therefore excluded this variable from further analysis, which thereby allowed us to expand our sample to include those cases where there were only nonmonetary recoveries.
A formal regression model (using ordinary least squares) was employed to explain the levels, alternatively, of the aggregate fee per case and the hourly rates awarded. The values of fees, target assets, and deals were expressed in terms of natural logarithms for the purposes of these regressions. These results are found in Table 9. As can be seen, a settlement that involved only nonmonetary relief yielded lower aggregate fees and lower hourly fees, consistent with the pattern found in Figure 9. In addition, aggregate fees and hourly fees tended to be higher where the target was larger (as measured by the target's assets) or where the deal value was larger. Finally, the hourly fees tended to be lower in settlements where there was a court-ordered reduction in fees (which is discussed at greater length in section F below), as expected, but the effect of this influence is not statistically significant.
Two other features of these legal fees are worth noting. First, the average size of the aggregate fees per case as a percentage of the claimed monetary recoveries achieved for the 31 settlements where there were such recoveries claimed was 4.6 percent, but the range was 0.01 percent to 29.6 percent, and the median was 1.9 percent. There was a significant negative correlation between the fee percentage and the size of claimed monetary recoveries. Second, the average size of the aggregate fees per case as a percentage of the deal value was 0.19 percent, with a range of 0.005 percent to 1.36 percent and a median of 0.12 percent. Thus, while plaintiffs' attorneys appear to have been well compensated, the attorneys' fees that they received increased by only a relatively trivial amount the total cost of these transactions to the acquiring corporations.
D. A More Detailed Analysis of Settled and Dismissed cases
In an initial effort to describe the pattern of settlements and dismissals-that is, to try to explain or predict which suits were likely to settle and which were likely to be dismissed, based on the characteristics of the challenged mergers in each category-we employed similar statistical analyses to those described above: examinations of means and medians of the merger characteristics and logit regressions. However, there were no significant differences between the characteristics of those challenged mergers where the lawsuits settled and those where the suit was dismissed.
We also conducted a more qualitative analysis of the pattern of settlements and dismissals against the backdrop of our predictions in Part II concerning the probability and probable course of litigation relating to the various categories of merger. To set the stage for this analysis, we provide below a somewhat more detailed and nuanced summary of the legal outcomes of the 104 instances in which mergers were challenged:
We will follow this organizational scheme in the discussion that follows.108
1. Mergers involving a conflict of interest
a. Mergers involving a controlling shareholder
Our prediction was that the typical pattern of such cases would be that the controlling shareholder would make an offer to squeeze out the public shareholders at a price lower than the highest price it was prepared to pay. A class action would be filed challenging the fairness of the proposed transaction. The controlled corporation would create an SNC, which would retain its own financial and legal advisors, provide plaintiffs' attorneys with an opportunity to make some sort of presentation in support of their view that the initial offer was unfair, and negotiate a higher price with the controlling shareholder. Plaintiffs' attorneys would acquiesce in the price agreed to by the SNC in an MOU in which defendants would acknowledge that plaintiffs' efforts contributed in some fashion to whatever increase in price the SNC had negotiated. Plaintiffs would also obtain the right to take confirmatory discovery and, perhaps, to comment on the proxy statement or tender offer documents that would be used to effect the squeeze out. Following such discovery, plaintiffs' attorneys and defendants would sign a formal Stipulation of Settlement, in which defendants would agree to pay plaintiffs' attorneys' fees and expenses, up to some agreed upon amount that would represent no more than a very modest percentage of the price increase (and a considerably more modest percentage of the value of the transaction), if awarded by the court.
The results of these cases were largely consistent with this prediction. All thirty-one cases involving challenges to squeeze outs were filed by members of the plaintiffs' bar. Of the eighteen cases settled for a monetary recovery, seventeen followed the above-described pattern very closely.
The only settled case that did not was a challenge to Citicorp's acquisition of the 15 percent of the stock of Travelers Property Casualty Corporation that it did not own.109 When Citicorp announced its initial offer of $41.50 cash per share, it also announced that the offer had been approved by an SNC of the Travelers board. This structuring of the merger effectively denied plaintiffs' attorneys any opportunity to free ride on the efforts of the SNC. However, eleven class-action complaints were filed challenging the fairness of the merger and alleging that the SNC was a sham. Moreover, in contrast to the other seventeen cases in this category, in none of which did plaintiffs' attorneys pursue a challenge to the independence of the SNC, here plaintiffs' attorneys filed a second Amended Complaint alleging that the Form 14D-9 filed by Travelers withheld and obscured material facts regarding the independence of the Chairman of the SNC and the fairness analysis prepared by Morgan Stanley for the SNC. Here alone plaintiffs' attorneys also sought expedited discovery and a preliminary injunction. That precipitated settlement discussions resulting in an agreement that Citicorp would increase its offer to $41.95 cash per share, which increased by $25.7 million the amount to be received by the plaintiff class.
In essence, when the plaintiffs' attorneys who challenged the Citicorp-Travelers merger could not free ride (because the Travelers' SNC had blessed Citicorp's bid before it was announced), they were prepared to engage in significant litigation efforts to challenge the SNC's decision. They then acted like true shareholder champions rather than merely self-interested litigators. That this was the only squeeze out case involving a monetary settlement in which plaintiffs' attorneys mounted such a challenge fueled our suspicions that in the other such cases, plaintiffs' attorneys' decisions to settle, rather than to challenge the improved prices negotiated by SNCs, may well have reflected their rational responses to the incentives provided by Delaware law, and not necessarily (or entirely) their good faith judgments that the terms of those mergers were "entirely fair."
None of the five settlements that involved only a non-monetary remedy followed the transaction/litigation pattern that we suggested would likely yield a monetary recovery. Consequently, they are not inconsistent with our self-interested litigator hypothesis. In three of the cases, the plaintiffs challenged the procedures that the defendants had followed in effecting the merger, and in the remaining two, an SNC had already been formed and had approved the merger price before the merger was announced. In four of the five cases, the plaintiffs settled for additional disclosures.110
There were eight cases in which challenges to squeeze outs were filed and then voluntarily dismissed without significant litigation by plaintiffs' attorneys. In three cases, the proposed transaction fell apart. In the fourth case, an SNC had already formed and approved the merger price prior to the merger announcement, and in the fifth, the defendants simply moved to dismiss and the plaintiffs failed to respond. There were three cases where an SNC was appointed and negotiated an improved price for shareholders, but where the plaintiffs' attorneys did not, or were somehow unable to, free ride on the price improvements achieved by the SNCs. Neither the litigation documents nor press reports provide facts that explain these outcomes, but one possible explanation is that the SNCs in these cases may not have given the plaintiffs' attorneys any opportunity to participate in their deliberations and thus may have denied them any clear basis to claim credit for the improvements in price. In any event, the fact that these SNCs negotiated improvements in the absence of active litigation provides some support for the argument that some (and perhaps most) SNCs are prepared to represent public shareholders' interests adequately without plaintiffs' attorneys looking over their shoulders.
b. Mergers involving other conflicts of interest
Our prediction in cases involving MBOs was very similar to that concerning squeeze outs. We also stated that it was more difficult to predict the patterns or outcomes of cases involving other conflicts of interest, other than that it seemed more likely that those cases would involve judicial decisions addressing the merits of plaintiffs' claims concerning the existence of a conflict of interests.
Especially if we exclude the three settlements involving monetary recoveries that were not prosecuted by members of the "traditional" plaintiffs' bar, our results differ significantly from our predictions. Only a small proportion of the cases prosecuted (or dismissed) by plaintiffs' attorneys resulted in monetary recoveriesfour of twenty-eight, or 14 percent-and a much higher proportion of those cases, 61 percent, were voluntarily dismissed. The disparity, we believe, can be explained largely by the fact that we anticipated that the transactional pattern of MBOs would be much the same as that of squeeze out mergers,111 but we found that not to have been the case. When a controlling shareholder proposes a squeeze out, it often will have no good reason to keep its bid a secret, since it has no duty to sell its shares and therefore no competing bid will be made. However, when an MBO is proposed, or when a board is considering strategic alternatives that may include an MBO, there may be good reason to believe that the highest value reasonably available can be obtained through private negotiations, rather than through a public "auction." We inferred that this often was the case in the period under study because a substantial proportion of the MBOs that we reviewed were publicly announced only after an SNC had tested the market and then agreed on terms with the MBO group. Although a plaintiffs' attorney still could challenge such a merger by alleging disclosure violations, a class action challenging the MBO price would be much less attractive because no increase in price above the price already negotiated by the SNC was likely, which would rule out the possibility of free riding on the efforts of the SNC. In addition, in MBO cases (as opposed to squeeze outs), an SNC often would find it relatively easy to prove that the price that it had negotiated was fair: it could point to its efforts to test the market to show that the price it had negotiated represented the highest value reasonably available for the target corporation's stock.112
Of the seven settled cases involving monetary recoveries, four were prosecuted by "traditional" plaintiffs' attorneys, and three were prosecuted by attorneys who represented plaintiffs with substantial financial stakes in the litigation. All of the four that were prosecuted by "traditional" plaintiffs' attorneys involved MBOs or mergers with similar characteristics. In three of them, consistent with our predictions, SNCs negotiated increased offers after the MBOs were announced, and in all three, plaintiffs' attorneys acquiesced in those offers. In the fourth, Seagate Technology,113 plaintiffs' attorneys obtained an order certifying the action as a class action, conducted extensive document discovery, took depositions, and moved for an expedited hearing on their motion for a preliminary injunction.114 Defendants then agreed to modify the terms of a somewhat complex MBO agreement, thereby providing $112.5 million to $200 million in additional value to the plaintiff class.115 In essence, as was true for Travelers (discussed above), where the plaintiffs' attorneys decided it was worth their while to invest in significant litigation activity, they were able to generate significant benefits for the plaintiff class. We find it noteworthy, though, that Seagate and Travelers are the only two cases in which "traditional" plaintiffs' attorneys' efforts clearly were responsible for producing significant monetary benefits for the plaintiff class. It also is noteworthy that in both cases the attorneys fees awarded were appreciably above the average-about 10 percent of the benefit produced in Seagate and 17 percent in Travelers.116 We will discuss the three cases that were prosecuted by attorneys representing plaintiffs with substantial stakes in Section E below.
Of the five cases that were settled for nonmonetary relief, three involved MBOs in which the price had been negotiated by an SNC before the merger was announced and there was no subsequent increase in price. A fourth involved an MBO in which the price was subsequently increased but where defendants did not credit plaintiffs for contributing to the increase. These four were settled for additional disclosures, revisions of deal protection terms, or various other forms of nonmonetary relief. In the fifth case, Captec Net Lease Realty,117 plaintiffs challenged the fairness of a side deal involving the sale of certain assets to the target's CEO.118 After the court granted plaintiffs' motion for expedited discovery, which suggests that it thought plaintiffs' claims had some merit, the parties settled for supplemental disclosure and payment by defendants of plaintiffs' attorneys' fees and expenses.119
Of the seventeen cases that were voluntarily dismissed, thirteen involved MBOs. In twelve of those, the terms of the merger had been negotiated by an SNC (or the target's full board in one case) prior to public announcement of the proposed merger. In the other, Authentic Fitness,120 the target's SNC negotiated a modest increase in price that plaintiffs' attorneys then challenged in an amended complaint.121 However, when defendants moved to dismiss, plaintiffs' attorneys did not file a response.122 They subsequently moved to dismiss without prejudice. In White Cap Industries,123 one of the twelve, plaintiffs' attorneys also filed an amended complaint challenging the independence of the members of the SNC and the fairness of the price.124 However, as in Authentic Fitness, when defendants filed and briefed a motion to dismiss, plaintiffs' attorneys did not respond on the merits and subsequently moved to dismiss without prejudice.
The other four dismissed cases involved allegations of other kinds of self-dealing in connection with mergers. In only one of them, BHC Communications,125 was there a significant level of litigation activity.
2. Mergers not involving a conflict of interest
a. Mergers clearly involving a sale of control
We predicted that challenges to mergers involving arm's-length sales of control were not likely to result in monetary recoveries, absent the emergence of a competing offer, but that complaints making such allegations would have some settlement value because they would allow plaintiffs' attorneys to impose on defendants the litigation costs involved in proving that they had met their Revlon duties and would also provide defendants with an opportunity to obtain "litigation insurance" at a relatively modest cost.
Our results were largely consistent with that prediction. In no such case did a competing bid emerge. Six (50 percent) of the cases settled, but none of those settlements involved any monetary recovery. Four of the settlements involved additional disclosures, none of which appeared to support plaintiffs' initial claims of unfairness. In the remaining two settlements, plaintiffs' attorneys negotiated provisions that had some potential to produce improved terms for public shareholders but that in reality yielded nothing.
Of the six (50 percent) cases that were voluntarily dismissed without prejudice, none involved a target for which a higher bid was made after suit was filed. In two, plaintiffs' attorneys moved to dismiss after defendants filed and briefed motions to dismiss.
b. Mergers arguably involving a sale of control.
We predicted that challenges to transactions in this category would be less attractive to plaintiffs' attorneys because they would likely be required to litigate on the merits the question of whether Revlon applied, but that they also would be more likely to result in monetary recovery if plaintiffs succeeded on that issue, since the target's board probably would not have sought the highest value reasonably available.
Only three mergers in this category were challenged. One case settled, and the other two were dismissed. The settled case, IXC Communications,126 was largely prosecuted by a former executive of IXC who held 2.4 percent of its stock, worth roughly $44 million.127 (We will discuss it further in section E below). In the two cases that were dismissed voluntarily, plaintiffs' attorneys claimed that the defendant corporations had agreed to sales of control without complying with Revlon. In neither case did a competing bid emerge, and in neither case did plaintiffs' attorneys engage in significant litigation activities after filing their complaints.
c. Mergers involving defensive tactics
We predicted that plaintiffs' attorneys would challenge mergers in this category either because a competing bid seemed likely (or was pending) or in the hope that a competing bid would emerge. If one did, we predicted that plaintiffs' attorneys probably would be in a position to realize substantial fees largely as a result of the efforts of the competing bidder. If no bid emerged, we predicted that they would probably abandon their claims by moving to dismiss without prejudice because success on the merits would be unlikely.
Our results largely were consistent with these predictions. Of the four class actions that resulted in monetary settlements, three challenged mergers or defensive tactics where a hostile bid had been made, and the fourth challenged a merger that was subsequently challenged by a hostile bidder. In all four cases, the hostile bidder assumed primary responsibility for challenging the target company's defensive tactics, and in all four cases the target company was acquired for a higher price. Nevertheless, plaintiffs' attorneys sought and received substantial fees of $1,050,000 ($1,085 per hour), $3,750,000 ($2,750 per hour), $9,450,000 ($1,630 per hour) and $625,000 (hourly information was not available), the last of which was paid by the acquirer after the original merger was abandoned and the target agreed to a second, friendly deal at a higher price.
In three of the eight cases that were dismissed, no higher bid emerged. Higher bids were made for the defendant corporations in three others, and an arguably higher bid was made for the defendant in a fourth. Further, in a fifth case, Newport News Shipbuilding,128 a hostile bid of equal value succeeded after the Department of Defense, whose support was critical, expressed its support for the second bidder.129 One plausible explanation for why plaintiffs' attorneys did not engage in more significant litigation activity in these four (or five) cases relates to the fact that in none of them did the successful bidder initiate a lawsuit challenging its target's allegedly improper defensive actions. This suggests that the bidders did not view those defenses as significant impediments, which also would tend to undercut any Unocal claims made by plaintiffs' attorneys. In addition, in the absence of lawsuits by the bidders, plaintiffs' attorneys had no opportunity to free ride on the litigation efforts of others.
d. Mergers not involving a sale of control or defensive tactics
Our predictions concerning these cases were similar to those concerning cases challenging mergers involving defensive tactics: plaintiffs' attorneys would file on spec, free ride if a competing bid emerged, and otherwise move to dismiss without prejudice because they would have little prospect of success on the merits.
Our results were entirely consistent with these predictions insofar as they related to the ten cases that were dismissed. None of the defendant corporations became the target of a competing bid, and in none of those cases was there significant litigation activity by plaintiffs' attorneys after the complaints were filed.
The two cases that were settled for monetary recoveries did not involve competing bids but nonetheless were not inconsistent with our predictions. Both grew out of post-announcement events that could not easily have been anticipated when the class actions originally were filed. It is noteworthy that in both cases the plaintiffs' attorneys accepted relatively modest fees.
3. Tender offers by a controlling shareholder
We predicted that cases challenging tender offers by controlling shareholders were more likely to result in, and effectively to be resolved by, substantive litigation because the law concerning such transactions was relatively unsettled for most of the period we studied. That proved to be the case.
Three such tender offers were challenged. In two, preliminary injunctions were sought and denied. Both cases were thereafter dismissed (although, in one case, the tender offer also failed). In the third case, plaintiffs' attorneys engaged in some litigation activity, including taking one deposition, but then moved to dismiss without prejudice.
E. cases with '"Real" Plaintiffs
In addition to the analyses described in Section D, we also examined class actions prosecuted by law firms that are not part of the "traditional" plaintiffs' bar130 on behalf of clients with substantial financial stakes in the outcomes of those actions.131 Our thought was that if attorneys representing "real" clients prosecuted class actions in much the same fashion as did "traditional" plaintiffs' attorneys, then that would suggest that "traditional" plaintiffs' attorneys were acting as faithful champions of shareholders' interests, while if the actions prosecuted by attorneys representing "real" clients differed significantly, then the differences would suggest that "traditional" plaintiffs' attorneys act more like self-interested litigators.
We found five cases that were actively prosecuted on behalf of clients with substantial financial interests by law firms not generally identified as part of the "traditional" plaintiffs' bar.132 In two cases, Digex133 and Telecorp PCS,134 such law firms acted as lead counsel for the plaintiff class. In two other cases, IXC Communications Corporation135 and SFX Entertainment Incorporated,136 such law firms were not appointed lead counsel for the class but were allowed to litigate on behalf of their clients, subject to a requirement that they coordinate their efforts with those of lead counsel for the plaintiff class. In the fifth case, Siliconix,137 such a law firm's client was appointed lead plaintiff, and the firm was appointed co-lead counsel, together with a firm generally identified as part of the "traditional" plaintiffs' bar.
The pattern of litigation in these five cases differed remarkably from that in the other cases we studied. In none of them did the "real" client file its complaint on the first or second day after the challenged merger was announced. The earliest complaint was filed six days later, the latest was filed thirteen days later, and the median interval between announcement and filing was nine days. This suggested to us that substantial consultation between these attorneys and their clients probably had occurred before the clients authorized the filing of a complaint. Moreover, the complaints themselves tended to be considerably more detailed than the "bare bones" complaints initially filed in these five actions or the complaints typically filed in other merger-related cases by the "traditional" plaintiffs' bar.
The nature of the claims asserted by these substantial shareholders also was very different from the merger-related claims most frequently asserted by "traditional" plaintiffs' attorneys challenges to the fairness of the price offered or agreed to in squeeze outs, MBOs, and arm's-length mergers. The substantial shareholders tended to make highly fact-specific claims that raised unique, rather than generic, legal or factual issues. Plaintiffs in Digex, SFX Entertainment, and Telecorp PCS alleged that defendants had engaged in unfair self-dealing on the basis of facts specific to each of those transactions.138 Plaintiff in IXC Communications made an unusual claim to the effect that IXC's board had agreed to sell the company in a stock-for-stock transaction solely for the purpose of avoiding its obligations under Revlon and also charged the acquirer with improperly paying cash, rather than stock, to another large shareholder to induce it to vote in support of the merger. Only Siliconix involved a challenge to the fairness of the price being offered, and that case involved a tender offer by a controlling shareholder-a setting in which, as noted above, considerable uncertainty existed as to the governing legal principles.
The attorneys representing all five of these substantial shareholder-plaintiffs engaged in extensive litigation activity following the filing of their complaints-actions consistent with a belief that the claims asserted had more than nuisance value. In three cases, Digex, IXC Communications, and Siliconix, the attorneys representing the plaintiffs engaged in substantial expedited discovery and then briefed and argued substantive motions for preliminary injunctive relief, succeeding in part in Digex but not in IXC Communications or Siliconix.139 In SFX Entertainment, the attorneys for the shareholder-plaintiff also conducted expedited discovery and then moved for summary judgment, precipitating a substantial settlement.140 In Telecorp PCS, the plaintiffs' attorneys successfully defended a motion to dismiss and thereafter engaged in substantial litigation activity in support of their claims. In contrast, in only three of the other ninety-nine cases that we studied did "traditional" plaintiffs' attorneys brief and argue similar substantive motions,141 and in only one case did a "traditional" plaintiffs' attorney succeed in whole or part with respect to such a motion.142
When the cases prosecuted on behalf of substantial plaintiffs were settled, the nature of the settlements and fee arrangements also differed substantially from the norm in cases prosecuted by the "traditional" plaintiffs' bar. Three of these cases resulted in very substantial recoveries-$165 million in Digex, $47.5 million in Telecorp PCS, and $34.5 million in SFX.143 Moreover, in all three of these cases the substantial plaintiffs attorneys claimed that the recoveries were due solely to their efforts144-a claim that was made in connection with only two of the twenty-seven settlements involving monetary recoveries that were negotiated by "traditional" plaintiffs' attorneys.145 In all three of these cases, too, the plaintiffs' attorneys' fee requests were consistent with their claim of sole credit, in that the fees requested represented a substantially higher percentage of the amounts recovered than did the fees requested by "traditional" plaintiffs' attorneys where the latter's litigation efforts were purported to be only partially responsible for whatever financial benefit was realized by the plaintiff class. Finally, in none of these three cases did the substantial plaintiffs attorneys attempt to negotiate a fee award payable by defendants. Rather, in all three the plaintiffs' attorneys sought judicial approval of fees payable from the common fund that their efforts had created. This contrasts sharply with the fee arrangements negotiated by "traditional" plaintiffs' attorneys, which, in all of the cases that we studied, involved agreements by defendants to pay (up to some agreed upon limit) whatever attorneys' fees the court awarded.
F. The Delaware Courts'Monitoring of Merger-Related Class Actions
Finally, we considered what the data that we had gathered suggested about whether the Delaware Chancery Court was effectively monitoring merger-related class actions to ensure that plaintiffs' attorneys were not exploiting their "license to litigate" primarily to enrich themselves. We did not give a great deal of weight to one datum-that all of the settlements negotiated by plaintiffs' attorneys received judicial approval-because courts in general are notoriously reluctant to disapprove negotiated settlements of complex corporate litigation.146 Nonetheless, the Chancery Court's 100 percent approval rate clearly provides no support for claims that the Court is acting as an effective monitor or is alert to the possibility of collusion.147
We found the reluctance of the Chancery Court to involve itself in the process by which lead counsel is appointed to be somewhat more troubling. We found that when multiple class-action complaints were filed, the court allowed (or encouraged) the plaintiffs' attorneys who filed complaints to resolve the lead counsel issue through private negotiations. This finding was consistent with the description that Franklin Balotti, an experienced Delaware practitioner, recently provided to the Third Circuit Task Force on Selection of Class Counsel.148 Indeed, as Mr. Balotti pointed out, "on the few occasions when Delaware courts have had to confront organizational disputes among plaintiffs' counsel, they have repeatedly admonished counsel to organize themselves."149
The plaintiffs' bar has responded to these admonishments by adopting a simple decision rule to determine which attorney or law firm should be appointed lead counsel: a presumption that the attorney or law firm who filed the first class-action complaint should be named lead or co-lead counsel.150 Thus, the Chancery Court's hands-off approach has had the unfortunate effect of promoting the "race to the courthouse" evidenced by our finding that the first complaints in the vast majority of merger-related cases were filed within one day after the challenged mergers were announced. Moreover, the court appears insensitive to the link between the approach that it has adopted and its expression of concern that "[t]oo often judges of this Court face complaints filed hastily, minutes or hours after a transaction is announced, based on snippets from the print or electronic media. Such pleadings are remarkable, but only because of the speed with which they are filed in reaction to an announced transaction."151
The findings to which we attach the most significance, though, are those relating to fee awards because we believe that the Chancery Court's authority to award attorney fees provides it with potentially its most effective mechanism to regulate the conduct of the plaintiffs' bar.152 We found forty-seven cases in which the Chancery Court passed on fee requests from "traditional" plaintiffs' attorneys.153 In forty of these, or 85 percent, the court awarded all fees requested. Two prior studies of Delaware Chancery Court litigation, one covering ninety-eight settlements of class and derivative actions reviewed between 1990 and 1992154 and another covering 138 settlements reviewed between January 1, 1998, and April 15, 2001,155 suggest that this is somewhat above the norm. Both of those studies found that all fees requested were approved in only two-thirds of the cases studied. Moreover, the absence of a demand requirement, which makes it considerably more likely that a class action will be filed and settled primarily because of its nuisance value, makes the higher rate of full approvals in merger-related class actions even more surprising.156
The specifics of the cases in which objections were filed and fees were reduced further suggest lax judicial oversight of the manner in which plaintiffs' attorneys litigated merger-related class actions. Table 10 lists the cases in which fees were reduced, the judge involved, whether an objection was made to the fee request of plaintiffs' attorneys,157 and the percentage by which the court reduced the fees requested.158 There were only two cases in which objections were filed. Both involved somewhat serendipitous circumstances. In Banctec, an objection was filed by Professor Weiss's colleague,159 and in Donna Karan International, one was filed by a class member that initially was a named plaintiff in that action and that frequently has been a named plaintiff in corporate litigation,160 but that apparently had a falling out with the plaintiffs' law firm with which it had been aligned.161
One could view the small number of objections as evidence that shareholders in general are satisfied with the efforts of plaintiffs' attorneys, but we do not construe it in that fashion.162 As noted above, every case settled by a "traditional" plaintiffs' attorney involved a provision in the Stipulation of Settlement to the effect that attorney fees (up to some agreed upon amount) would be paid by defendants, rather than from a common fund or by members of the plaintiff class. By allowing settlements to be so structured, the Chancery Court has reduced considerably the likelihood that objections will be filed. Defendants in all such cases had committed themselves not to object to plaintiffs' attorneys' fee requests.163 And members of the plaintiff class had no financial incentive to object, since the fees were not coming directly out of their pockets-as would have been the case had those fees been payable from a common fund.164 Consequently, the Chancery Court's tolerance of this settlement structure itself suggested to us that the court had little interest in encouraging class members to scrutinize closely the fee awards requested by their self-appointed champions.
Even more striking, though, were the court's reactions to the arguments advanced by the two objectors. In Banctec, the objector argued that no fee should be awarded primarily because plaintiffs had filed a suit for which they had no factual basis, had found no evidence to support their claims, and had then negotiated a settlement calling for supplemental disclosures that provided members of the plaintiff class with information of no significance.165 In Donna Karan International,166 the objector argued that the suit had been filed prematurely because plaintiffs' attorneys should have waited to see if the board of Donna Karan International (DKI) accepted the challenged buy-out offer; that DKI was not controlled by persons affiliated with the offeror because another shareholder group owned almost as many shares; that the increase in price was attributable almost entirely to the efforts of DKI's SNC and the other large shareholder group; that the number of hours claimed by plaintiffs' attorneys was excessive given the limited litigation activity they had undertaken; and that most work claimed by plaintiffs' attorneys was done by partners even though it was routine.167
The court appeared to credit both objectors' arguments, in that it reduced plaintiffs' attorneys' fee request by 90 percent in Banctec and by 78 percent in DKI.168 Yet arguments identical or similar to those made by objectors in Banctec and DKI, in our view, could be made with comparable force with respect to a substantial majority of the class-action settlements that we examined. However, where no objector appeared to make them, the Chancery Court rarely took the initiative. In only one other case, IBP, did the court reduce the requested attorneys' fees by more than 50 percent, and there the benefit for which plaintiffs' attorneys claimed credit was, in a sense, antithetical to the theory on the basis on which those class actions initially had been filed.169 In only four other cases were fees reduced at all, and in all but one of those, the reductions were made by the same Vice-Chancellor.170 In 89 percent of the cases in which no objection was filed, the court awarded to plaintiffs' attorneys all of the fees that they requested.
The Chancery Court appears to base its hands-off approach on the fact that Delaware practice calls for plaintiffs' attorneys to refrain from discussing fees with defendants until after the substantive terms of a settlement have been agreed upon and that defendants consequently have "a 100 percent direct dollar-for-dollar interest in getting the lowest fee [they can]."171 This reasoning, it seems to us, reflects a naive or simplistic view of the dynamics of the settlement process.172 Attorneys for both plaintiffs and defendants in these cases almost always are repeat players. As such, they know at the time that they negotiate a MOU setting forth the terms of a proposed classaction settlement that the attorney fee issue ultimately will have to be addressed.173 They also know that if the proposed merger closes before confirmatory discovery is completed and the formal Stipulation of Settlement is signed-which typically is the case-plaintiffs' attorneys almost certainly will expect defendants to pay whatever fees subsequently are awarded.174 Moreover, as an economic matter, we find it hard to believe that defendants in these cases do not have some "bottom line" on the total amount that they are prepared to pay, to the plaintiff class and/or to plaintiffs' attorneys, in order to consummate a proposed merger. Thus, we suspect, some "shadow negotiation" of fee awards inevitably occurs. The vehicle for this negotiation may well be the conversation in which attorneys for plaintiffs and defendants discuss the wording of the clause in the standard MOU in which defendants acknowledge the extent to which plaintiffs' attorneys' efforts have contributed to whatever benefit the class is purported to have realized.175
At a minimum, an examination of the settlements we studied, of the Chancery Court's decisions in the two cases in which objections were filed, and of that court's decisions in the much larger number of cases in which no objection was filed, suggests to us that the court did not use its power to control fee awards to monitor effectively the litigation efforts of the plaintiffs' bar. Why, for example, was the court prepared to reduce dramatically the attorneys' fees requested in Banctec and Donna Karan International, but not equally prepared to make similar cuts in the fees requested in the far larger number of cases in which similar objections could be made but no objector appeared? That the defendants in those other cases were prepared to pay the fees requested does not strike us as a persuasive response, for the reasons outlined above.176 What the court should have appreciated, we believe, is that defendants' willingness to pay plaintiffs' attorneys' fees may well have represented either a strategic concession that so proceeding was either the least costly means of disposing of a lawsuit that lacked merit but had nuisance value or an implicit payoff to the plaintiffs' attorneys in exchange for their acquiescence in the terms negotiated by an SNC. By approving those fee arrangements-which usually provided plaintiffs' attorneys with very high hourly compensation177-the Court effectively encouraged plaintiffs' attorneys to continue litigating merger-related class actions in a manner that, by and large, appears to advance only the interests of those attorneys.178
G. An Assessment
We began this Part by laying out two alternative hypotheses as to what drives shareholder litigation, especially with respect to mergerrelated class actions in the Delaware courts: the shareholder champion hypothesis and the self-interested litigator hypothesis. We believe that the evidence presented in sections B-F, taken as a whole, strongly supports the self-interested litigator hypothesis.
In Section B we provided an analysis of which mergers tend to be challenged by class-action shareholder lawsuits. We found that important influences on the likelihood that plaintiffs' attorneys would challenge a merger were the size of the transaction, the (inverse of the) size of the initial premium offered to the target company's shareholders, the presence of an all-cash deal (representing a sale of control), and the presence of an acquiror with a significant prior stake in the target or other appearances of conflict of interest. Though these influences could be interpreted as consistent with either hypothesis, the overall conclusion nevertheless is that this particular type of litigation is systematic and purposive and not the result of random decisions on the part of plaintiffs or their attorneys.
The tendency of mergers to be challenged where the acquiror had a prior stake or there were other conflicts poses an especially interesting and subtle problem of interpretation. It may be the case that there are continuing abuses in these kinds of transactions (although it is unclear why the prospects of costly litigation have not deterred them). But another possibility should be considered. Could it be that, given the structure of Delaware law that we described in Part II, an acquiror's prior stake in the target (or an acquiror's arrangements to continue to employ and reward the incumbent management), and the possibility of insider abuses that are thereby raised, often presents too tempting a target for plaintiffs' attorneys? If this is so, and especially if an SNC has not yet had a chance to consider the acquiror's offer at the time that the offer is made public, then an acquiror with a prior stake may simply have to face the likelihood that a class-action lawsuit will be filed, regardless of the price that the acquiror initially offers to the other stockholders. This prospect would surely reinforce the bargaining instincts that in any event might lead an acquiror initially to make a "lowball" offer, which it then would be prepared to increase in subsequent negotiations with the SNC. Although such lowballing would increase the probability that a class action will be filed following the announcement of a squeeze out or an MBO, thus generating an unfortunate equilibrium of lawsuits' being filed whenever acquirors have prior stakes in their targets, it also might present acquirors with opportunities to insulate such mergers from scrutiny by arranging advantageous settlements with the plaintiffs' bar.179 The creation of this kind of equilibrium would be most consistent with the self-interested litigator hypothesis.
We next turned in Section C to an overview description of the 104 mergers that were challenged and the characteristics of the litigation. An initial complaint was filed within the first two days after the merger announcement in over 80 percent of the merger challenges. Multiple complaints were filed in most cases. This pattern of early and frequent filings is one that is consistent with opportunistic behavior by plaintiffs' attorneys-specifically, the self-interested litigator hypothesis.
Settlements occurred in slightly less than 50 percent of the cases that have been resolved. No cases were litigated to a judgment on the merits in favor of the plaintiffs. Of the cases that were dismissed, over 90 percent were dropped before there were any judicial decisions that addressed the merits of the lawsuit, and very few of these voluntary dismissals involved significant litigation efforts by plaintiffs' attorneys. Where the settlements involved claimed monetary recoveries, the plaintiffs' attorneys most often appeared to have free ridden on the decisions of SNCs or on improved terms offered by a competing bidder, rarely claimed full credit for the purported recovery, and never persisted in challenging improved terms negotiated by an SNC (or the terms proposed by a competing bidder). Further, in settled cases, although plaintiffs' attorneys' requests for fees tended to be a low percentage of the claimed recoveries180 (arguably an admission by the attorneys that they contributed little to the recoveries), they generally represented rich rewards when measured on an hourly basis. Plaintiffs' attorneys also tended to do well even when settlements did not involve any monetary recovery. Since settlements occurred in slightly less than half of the lawsuits filed, and since the hours and efforts devoted to the lawsuits that were dismissed generally were very modest, the overall returns to plaintiffs' attorneys from filing merger-related class actions appear to be attractive. Again, this pattern is consistent with opportunistic behavior by plaintiffs' attorneys.
The fact that the aggregate legal fees in settled cases tend to be a relatively low percentage of claimed monetary recoveries has two potential interpretations, the second of which adds to our sense of unease about this process. On the one hand, lower fees could be evidence of the efficiency (low relative transactions costs) and the modest aspirations of the plaintiffs' bar (and the reason why Delaware courts only infrequently award less than all fees requested). On the other hand, they can be seen as a practical admission that plaintiffs' attorneys' efforts contributed little to the improved terms in these transactions and that plaintiffs' attorneys were largely free riding on improvements that would have occurred anyway. We find the second interpretation more persuasive.
We next provided in Section D a summary of our detailed examination of all 104 challenged mergers and the accompanying litigation against the backdrop of the type of merger. This analysis again supports a picture of free riding with little "value added" by plaintiffs' attorneys in most cases, again supporting the self-interested litigator hypothesis.
We recognize that our analysis in Section D does not answer definitively one key question: whether the class actions that we studied and, in particular, the cases involving improvements in the terms of squeeze outs and MBOs, either (a) actually produced improvements in the terms of mergers; (b) involved no more than free riding by plaintiffs' attorneys; or (c) often involved tacitly collusive settlements in which plaintiffs' attorneys dropped their objections to arguably unfair mergers in exchange for defendants' implicit assurances that they would finance relatively generous attorney fee awards.181 It may be, as members of the plaintiffs' bar have argued to us, that plaintiffs' attorneys are so committed to vigorous advocacy of shareholders' interests that the incentives provided by Delaware law, even if perverse, are not sufficient to cause them to deviate from that commitment. However, we remain skeptical that such is the case. Our other findings (and economic theory) all suggest it is far more likely that plaintiffs' attorneys are motivated primarily by self-interest and that their litigation efforts, shaped as they are by the incentives provided by Delaware law, produce little in the way of meaningful benefits for the shareholders that those attorneys purport to represent.
In Section E we contrasted five cases, where "real" plaintiffs with substantial stakes were represented by attorneys who were outside the "traditional" plaintiffs' bar, with the remaining ninety-nine cases that were prosecuted by the traditional bar. In those five cases, litigation was more substantial; in three, the plaintiffs achieved significant monetary recoveries; and, reflecting the significant roles that they played, plaintiffs' attorneys in those three settlements received fees that were a higher fraction of the recoveries than was the norm and that were paid out of a common fund rather than directly by defendants (as was the norm in cases settled by "traditional" plaintiffs' attorneys). This sharply different litigation pattern for these five cases suggests to us that the norm for the other ninety-nine is better explained by the self-interested litigator hypothesis.
Finally, in Section F we examined the Delaware courts' reviews of settlements and of attorney fee awards. We found that no settlements were overturned by the courts, and that in 85 percent of the settled cases the courts approved the fee requests from the "traditional" plaintiffs' attorneys; further, the fees are almost always paid by defendants rather than out of a common fund, so that members of the plaintiff class have little basis for objecting. We argue that this record of relatively lax review is likely to encourage a pattern of litigation driven by the self-interested litigator hypothesis.
In sum, the pattern that we observe in the class-action shareholder lawsuits that were filed against merger targets in Delaware courts is redolent of a pattern of opportunistic filings, of a lawyer-driven process rather than a true client-driven process:
* systematic behavior with respect to which mergers were challenged;
* early and frequent complaints filed;
* a very high percentage of dismissed cases never reached a judgment on the merits;
* the absence of a single case that has been decided in favor of the plaintiffs on the merits;
* settlements tending to reflect free riding by plaintiffs' attorneys;
* plaintiffs' attorneys failing to persist in challenges to SNCs' decisions or to competing offers;
* attorneys with "real" clients and from outside the "traditional" Delaware plaintiffs' bar were far more vigorous in their litigation efforts on behalf of their clients;
* no settlements overturned by the Delaware courts;
* plaintiffs' attorneys' fee awards in settlements usually paid by defendants and not out of common funds, and largely unchallenged; and
* plaintiffs' attorneys' fees representing a strikingly low percentage of claimed recoveries (even though they are attractive on an hourly basis), which may well indicate that the attorneys added little value to the recoveries.
In sum, it is a pattern that is far more consistent with the self-interested litigator hypothesis than with the shareholder champion hypothesis. We find this picture quite troubling, as it suggests that merger-related class actions primarily serve as a vehicle through which the plaintiffs' bar (and to some extent the defense bar) are able to extract rents from the shareholders of Delaware corporations.
We began by hypothesizing that Delaware law provides plaintiffs' attorneys with substantial incentives to initiate, prosecute and settle merger-related class actions in a manner that advances their own economic interests rather than those of the shareholders they purport to represent. Our analysis of relevant economic and litigation-related data strongly suggests that the plaintiffs' bar, at least with respect to mergers announced in 1999 through 2001, responded to these incentives in a predictably self-interested fashion. We found those results troubling, especially since the litigation environment in Delaware today is largely the same as it was during the years that we studied.
Merger-related class actions have become the dominant form of shareholder litigation in Delaware. If such actions continue to serve primarily as devices that enrich plaintiffs' attorneys while providing shareholders with little in the way of meaningful benefits, they will undermine an important pillar of Delaware's system of corporate governance. As Professor James Cox has observed:
Much like the shepherd who cries wolf too frequently, shareholder suits, if commonly understood to be frivolous, will not in their commencement, prosecution and settlement affirm the social norms the suit's defendants allegedly violated. Their defendants will instead be seen as the objects of bad luck not derision. Thus, the procedural context in which corporate and securities norms are developed and affirmed are of the utmost significance if those norms are to discipline managers.182
We are confident that Delaware's courts did not deliberately set out to construct a litigation environment with such a strong propensity to generate agency costs. Rather, the current, unsatisfactory situation appears to have evolved largely as a consequence of the interaction between Delaware's existing system of privatized enforcement and a series of recent decisions that address, seriatim, a variety of difficult substantive problems relating to mergers, takeovers, and takeover defenses. Nonetheless, given the unfortunate impact that those substantive decisions appear to have had on merger-related class-action litigation, some remedial action by Delaware's courts appears to be in order.
Specifying the remedial actions the courts should take, however, poses a ticklish problem. Corporate governance and corporate litigation involve dynamic processes. How they will be affected by any given change in the law often is hard to predict and, as data in this paper make clear, well-meaning changes often produce unintended (and unfortunate) consequences.183 Consequently, we set forth our suggestions for change with a strong sense of modesty. We recognize that, as academics, we may not be sufficiently sensitive to practical problems that those changes may produce for the courts or to how either plaintiffs' or defense attorneys are likely to react if the changes that we propose are implemented.184
Our first suggestion relates to the process by which lead plaintiff and lead counsel are appointed. Given the time pressures that surround most merger-related class actions, we do not believe it feasible for Delaware courts to implement procedures similar to those used by federal courts to appoint lead plaintiffs and lead counsel in securities class actions. However, we believe it would be both feasible and desirable for the Chancery Court to signal that it is not necessarily antagonistic to efforts to involve it in the lead plaintiff/lead counsel appointment process, so as to encourage institutional investors and other substantial shareholders to become more actively involved in merger-related class actions.185
Absent judicial intervention, the "traditional" plaintiffs' bar is highly unlikely to relinquish voluntarily to attorneys representing "real" clients186 the control it currently exercises over most merger-related class actions. Moreover, if the Chancery Court continues to signal to institutional investors that it usually will react with annoyance to efforts on their part to enlist its help in obtaining control of merger-related class actions, institutions are likely to decide that it doesn't pay even to make such efforts. To avoid this outcome, when future cases arise in which institutional investors seek to become lead plaintiffs and to have their attorneys appointed lead counsel, the court should clarify that whatever antagonism it has expressed toward efforts to involve it at this stage of litigation pertains only to internecine spats within the plaintiffs' bar-that it does not extend to efforts by attorneys with "real" clients, proceeding on the basis of well-researched pleadings, to gain control of class actions from plaintiffs' attorneys whose claims to control are based largely on the fact that they were the first to file a class-action complaint.
Our second suggestion concerns pleading standards. In their most recent paper, Professors Thompson and Thomas conclude that derivative suits, when filed on behalf of shareholders of public corporations, generate relatively few litigation-related agency costs, in large part because the demand requirement and the possibility of special litigation committee review discourage opportunistic litigation.187 As we point out above, Delaware law imposes no comparable constraints on merger-related class actions, and, in part as a consequence, opportunistic class-action litigation appears to be common. Class-action complaints frequently are filed for strictly tactical reasons-almost half of the complaints we studied were dismissed voluntarily without the attorneys who filed them undertaking any significant additional litigation efforts-and very few of the remaining complaints ultimately lead to the realization of significant benefits by the shareholders on whose behalf they ostensibly are filed.
The Delaware Supreme Court may have sound reasons to continue to impose on corporate directors and controlling shareholders the burden of proving, with respect to a variety of merger-related decisions, that they proceeded in conformity with the relevant standard of review. But, it seems to us, so assigning the burden of proof does not necessarily require the court to rule, as it did in Krasner v. Moffett, that a bare bones complaint188 should not be dismissed.189 Rather, we believe that in cases where defendants have the burden of proof, the court nonetheless should dismiss class-action complaints unless plaintiffs plead particularized facts that, if true, would create a reasonable doubt that defendants failed to satisfy (or are reasonably likely not to satisfy) the relevant standard of review.190 In addition, the court should consider further discouraging the filing of frivolous or entirely speculative class-action complaints by encouraging imposition of sanctions on attorneys whose complaints clearly fall short of this standard or on attorneys who file complaints that they then seek to dismiss voluntarily without making any significant efforts to prosecute the claims advanced therein.191
Finally, and most importantly, Delaware courts should begin to review far more rigorously requests for attorney fee awards. It is the prospect of fee awards, Delaware courts have acknowledged, that motivates the plaintiffs' bar. Now those courts must recognize that it is essential that they use their power over fee awards to motivate the plaintiffs' bar to reshape its class-action litigation efforts.
In Parts II and III of this paper we lay bare-convincingly, we believe-the litigation dynamics that appear to underlie the vast majority of settlements of merger-related class actions. Most monetary settlements appear to involve free riding at best and tacit collusion at worst. All but a few nonmonetary settlements appear to result from plaintiffs' attorneys' ability to threaten to impose litigation costs on defendants. Yet Delaware courts, by and large, appear to have ignored the litigation dynamics that gave rise to these settlements. Instead, in case after case, those courts have tended to rubber-stamp whatever fee requests were made by plaintiffs' attorneys, so long as defendants did not object.192 By so proceeding, those courts have failed to fulfill the monitoring role that their judicially-created system of privatized enforcement assumes they will play.
At least two changes seem to us to be essential. First, the Delaware Supreme Court should reconsider its decisions in Tandycrafts and United Vanguard Fund. The presumption that they create-that any benefits realized by the plaintiff class after a class action has been filed were caused, at least in part, by the litigation efforts of plaintiffs' attorneys-appears to us to be counter-factual, especially as it applies to cases in which plaintiffs' attorneys have an opportunity to free ride on the efforts of others. In addition, that presumption and the related burden of proof193 increase dramatically plaintiffs' attorneys' leverage when bargaining for an attorneys' fee award in cases in which they can free ride on the efforts of others and, as such, also may promote settlements that are tacitly collusive. Perhaps most importantly, they limit dramatically the Chancery Court's ability to deny fees in such cases.
The best solution would be for the Supreme Court to eliminate the presumption in favor of a fee award whenever the class realizes some postfiling benefit.194 At a minimum, the Court should emphasize that its holdings in Tandycrafts and United Vanguard Fund are subject to its prior holding in Chrysler Corporation v. Dann to the effect that before a plaintiffs' attorney is entitled to a fee award, she must demonstrate that at the time she filed her complaint, she "possesse[d] knowledge of provable facts which h[e]ld out some reasonable likelihood of ultimate success."195 In many cases-particularly those in which the initial complaint was filed before the terms of a squeeze out or MBO have been finalized-strict enforcement of this standard will reduce considerably the ability of plaintiffs' attorneys to demand unjustified fee awards.
Second, when it reviews settlements and requests for attorneys' fees, the Chancery Court should take account of the litigation dynamics described above. Where plaintiffs' attorneys appear largely to have free ridden on the efforts of others or to have succeeded primarily because they were able to threaten to impose unnecessary litigation costs on defendants, any fee awards should be parsimonious.196
So changing the manner in which the Chancery Court reviews settlements and fee requests no doubt initially would add to its already heavy workload. But we believe that such an effort nonetheless would prove to be worthwhile. A series of parsimonious fee awards would send a clear and important message to the plaintiffs' bar-free riding and prosecution of nuisance litigation no longer will be profitable. Plaintiffs' attorneys, if they want to continue to receive the generous fee awards to which they have become accustomed, will have to change their approach to merger-related litigation. In the future, they will have to shift their focus to identifying mergers in which shareholders really are not being treated fairly and then litigating vigorously to protect those shareholders' interests. Merger-related class actions probably would become far less common, but those that are filed and prosecuted then are likely to produce far more in the way of genuine protection of shareholders' interests. In short, the benefits envisioned by the Delaware courts when they decided to privatize enforcement of merger-related fiduciary duties might actually be realized.
57 Vand. L. Rev. 1797 (2004)
Elliott J. Weiss*
Lawrence J. White**
* Charles E. Ares Professor, James E. Rogers College of Law, University of Arizona
** Arthur E. Imperatore Professor, Stern School of Business, New York University
We would like to thank Larry Cunningham, Lawrence Hamermesh, Alan Palmiter, Leo Strine and participants in the Vanderbilt/Institute for Law and Economic Policy conference for valuable comments, Virtual Docket LLC for providing us with discounted access to materials on its web site, and Guillermo Caravel, Anjela Kniazeva, Diana Kniazeva, Matthew Meaker, and Noel Mpaka for their research assistance on this paper.…