In Re Oracle Securities Litigation

852 F.Supp. 1437 (1994)

In re ORACLE SECURITIES LITIGATION.
This Document Relates to: All Actions

No. C-90-0931-VRW.

United States District Court, N.D. California.

May 24, 1994. As Amended June 14, 1994.

Richard B. Dannenberg, Lowey Dannenberg, Bemporad & Selinger, P.C., New York City, for class plaintiffs.

Andrew J. Ogilvie, Law Office of Andrew J. Ogilvie, Elizabeth J. Cabraser, Lieff Cabraser & Heimann, Stuart H. Savett, Kohn Savett Klein & Graf, Philadelphia, PA, for derivative plaintiffs.

Melvin R. Goldman, Darryl P. Rains, Morrison & Foerster, San Francisco, CA, Raymond L. Ocampo, Brenda G. Woodson, Oracle Systems Corp., Redwood Shores, CA, Lawrence M. Popofsky, Michael Rugen, Heller, Ehrman, White & McAuliffe, Frederick S. Fields, Stan G. Roman, Richard P. Walker, Bronson, Bronson & McKinnon, San Francisco, CA, for defendants.

Lawrence W. Schonbrun, Berkeley, CA, for class member David Teeters.

ORDER APPROVING CLASS AND DERIVATIVE SETTLEMENTS AND GRANTING ATTORNEY FEES AND COSTS.

WALKER, District Judge.

On the basis of the report of Oracle's special litigation committee, the settlement of these consolidated class and derivative actions can now be approved. This order explains the reasons for this approval and then addresses the fees and costs which derivative and class counsel should be awarded as a result of their efforts in this litigation.

I

Following the court's order dated August 9, 1993, In re Oracle Securities Litig., 829 F.Supp. 1176 (N.D.Cal.1993) ("Oracle IV"),^[1] Oracle created a special litigation committee ("SLC") made up of two "disinterested" directors, Joseph B. Costello and Delbert W. Yocam, to investigate and decide what action Oracle should take regarding the derivative litigation. By resolution, Oracle's Board of Directors ("Board") authorized the SLC to exercise all power and authority of the Board with regard to the derivative litigation. To assist in the investigation, the committee retained the law firm of Latham & Watkins as independent counsel.

The SLC presented a 67-page report to the court detailing its investigation. In its report, the committee concluded that it is in the best interests of Oracle to settle the derivative litigation under the terms worked out by Oracle's litigation counsel and that it is in the best interests of Oracle's shareholders to proceed with the settlement. Accordingly, the parties now re-petition the court for approval of the class and derivative settlements.

The pertinent facts and procedural history of this action are detailed in Oracle IV, 829 F.Supp. at 1177-78. In that order, the court scrutinized the class settlement in detail and found the terms of the class settlement reasonable. Id. at 1178-83. The court withheld approval of the class settlement solely because the class and derivative settlements were contingent upon one another and the court could not approve the derivative settlement due to the lack of independence of the directors and counsel making the decision to settle the derivative litigation. Id. at 1183-90. Because the court concludes that the SLC has acted with the independence required by Delaware law^[2] and that its conclusions with respect to the derivative settlement satisfy the business judgment rule, the proposed derivative settlement should now be approved. See Zapata v. Maldonado, 430 A.2d 779, 788-89 (Del.1981).

A

Whether a special litigation committee's decision to terminate derivative litigation is independent or not depends on the "totality of the circumstances." See Johnson v. Hui, 811 F.Supp. 479, 486 (N.D.Cal.1991) (applying Delaware law). Under Delaware law, "a director is independent when he is in a position to base his decision on the merits of the issue rather than being governed by extraneous considerations and influences." Kaplan v. Wyatt, 499 A.2d 1184, 1189 (Del. 1985). Delaware courts generally look to a number of factors when applying the "totality of the circumstances" test to a special litigation committee: (1) a committee member's status as a defendant, and potential liability; (2) a committee member's participation in or approval of the alleged wrongdoing; (3) a committee member's past or present business dealings with the corporation; (4) a committee member's past or present business or social dealings with individual defendants; (5) the number of directors on the committee; and (6) the "structural bias" of the committee. Id (citations omitted).

Both members of the SLC are "disinterested" directors of Oracle in the sense that they are in a position to base their decisions on the merits of the issues rather than extraneous considerations or influences. Delbert Yocam had no direct contacts or affiliations with Oracle or the defendants prior to becoming a director of Oracle. Since becoming a director, Yocam has had no contacts or affiliation with Oracle or the individual defendants outside his capacity as a director and shareholder. Moreover, Yocam was not a member of Oracle's board of directors at any time during the period in which the alleged wrongdoing involved in these cases occurred. From 1979 to 1989, Yocam held a number of executive management positions at Apple Computer, Inc., culminating with his service as executive vice president and chief operating officer. He is now president and chief operating officer of Tektronix, Inc., and is a director of Adobe Systems, Inc., and AST Research, Inc., all prominent firms in the computer industry.

Joseph Costello is president, chief executive officer and a director of Cadence Design Systems, Inc. ("Cadence"). He served as president and chief operating officer of SDA Systems, a predecessor to Cadence. Unlike Yocam, however, Costello has extensive prior contacts with some of the defendants and Oracle. The committee report summarizes these contacts:

Lawrence Ellison, one of the Defendants, was formerly a Director of SDA Systems. [Defendant] Donald Lucas was formerly Chairman of and a Director of SDA Systems. Mr. Costello met Mr. Lucas prior to becoming President of SDA Systems. Mr. Lucas interviewed Mr. Costello for his first position in SDA Systems as Director of Technical Operations. Mr. Costello was eventually promoted to Chief Executive Officer and was appointed to the board of SDA Systems during the same time that Mr. Lucas and Mr. Ellison served together on the SDA Systems board. Mr. Ellison nominated Mr. Costello to the Oracle Board. Mr. Costello met [defendant] Jeffrey Walker prior to joining the Board of Oracle but did not become acquainted with him until joining the Oracle Board. After Mr. Walker left Oracle, Mr. Costello and Mr. Walker purchased a piece of real property together in Jackson, Wyoming as an investment. Most of the property has since been subdivided. Mr. Costello has had no other contacts or affiliations with any of the Defendants prior to joining the Board.

SLC Rep at 22-23. Since joining the Oracle board of directors, however, Costello has had no contacts or affiliation with Oracle other than in his capacity as a director and shareholder.

Costello's contacts with the individual defendants admittedly implicate some of the factors that courts have used to identify lack of independence. Costello clearly has past business dealings with defendants Ellison, Lucas and Walker and these may arguably predispose him to a position favoring their interests.

A "totality of the circumstances" test does not, however, necessitate the complete absence of any facts which might point to non-objectivity. In any business setting, associations and contacts of the type which Costello has had with some of the individual defendants and Oracle are certainly neither inappropriate nor such as to suggest that Costello would not faithfully discharge his obligations to Oracle's shareholders. Business dealings seldom take place between complete strangers and it would be a strained and artificial rule which required a director to be unacquainted or uninvolved with fellow directors in order to be regarded as independent. See Johnson, 811 F.Supp. at 487. Costello's contacts, alone, do not demonstrate an interest or bias that would compromise Costello's objectivity.

Even, if Costello's background suggested some alleged interest, however, there is nothing to indicate that the SLC's judgment was tainted in any way. Costello was not the only member of the SLC and there is no evidence that Yocam's objectivity was affected by Costello's participation. See Johnson, 811 F.Supp. at 486 ("[t]he larger the number of directors, the less weight [is] accorded to any disabling interest affecting only one director").

In addition, the SLC retained Latham & Watkins as independent outside counsel. Neither Latham & Watkins nor any of the attorneys specifically assigned to advise the committee has had any previous connection or relationship with Oracle or any of the individual defendants. See Johnson, 811 F.Supp. at 487. The SLC's independent counsel, and their obvious interest in their reputation, acted as a further safeguard against any improper influence which Costello's associations with the individual defendants and Oracle might exert.

There is nothing before the court to indicate that the SLC conducted its investigation improperly. Before drafting its report and drawing its conclusions, the SLC reviewed large numbers of pleadings, transcripts, depositions, filings, corporate papers and other documents. The committee directed its counsel to conduct extensive interviews of counsel for the parties. Notes of these interviews were reviewed by the committee. Finally, the SLC reviewed all available materials in connection with the SEC investigation of Oracle involving most of the issues alleged in the derivative action.

In sum, the court finds that the SLC exhibited independent judgment and acted in good faith in conducting its investigation and arriving at its conclusions.

B

Having found the SLC and its counsel acted independently and in good faith, the court must now determine, under the business judgment standard, whether the SLC's conclusion that the derivative litigation should be settled on the terms negotiated by derivative plaintiffs and Oracle's litigation counsel satisfies the business judgment test. The court finds that it does.

In its report, the SLC concluded that it was in the best interests of Oracle to settle the consolidated derivative litigation rather than litigate further. It also found the terms of the proposed derivative settlement preferable to the available alternatives and recommended settlement on that basis. The SLC's reasoning is clearly detailed in its report. Only a brief summary of the SLC's rationale need be reprised here.

There are essentially two alternatives to the proposed settlement which the SLC considered: (1) continuing the derivative litigation, and (2) moving the court to terminate the litigation.

The first alternative, continuing the derivative litigation, is of course unappealing. Derivative litigation is unavoidably distracting to corporate officials whose efforts might more profitably be devoted to the company's business. Litigation makes sense only if the prospects of recovery are significant. The prospect of the derivative litigation producing any significant recovery for Oracle is indeed a dim one for several reasons.

Each derivative defendant has a written indemnity agreement with Oracle which indemnifies for any act "in good faith" or "reasonably believed to be in or not opposed to the best interests of the Company." SLC Rep, Ex 11, ¶ 4(b). Excluded from the indemnity provisions is "willful misconduct of a culpable nature." Id. The indemnity agreements require Oracle to advance to the individual defendants their costs of suits which the defendants are not required to reimburse unless they are found not to be entitled to indemnification. Id at ¶ 6. This means, of course, that Oracle faces the prospect of financing both sides of the derivative litigation.

A second obstacle to recovery lies in Oracle's restated certificate of incorporation. That document absolves Oracle directors of liability to the company or its shareholders for breach of duty as a director to the fullest extent permitted under Delaware law, limiting liability except for breaches of the duty of loyalty, bad faith, intentional misconduct or knowing violation of the law and the like. Id, Ex 12 at 2-3.

Needless to say, these provisions create forbidding obstacles to the prospects of recovery in the derivative litigation. Moreover, establishing the individual defendants' liability, even apart from these added problems, is highly uncertain on the record before the court.

Plaintiffs' insider trading claims have little support in the record. First, the SEC conducted a two-year investigation focusing in part on insider trading and did not charge the defendants with any violations of federal insider trading law. Second, all questioned sales of securities by defendants were during "window periods" authorized by Oracle's insider trading policy. Third, none of the questioned sales occurred anywhere near Oracle's peak stock price. Fourth, the amount of stock sold by defendants was a relatively small fraction of their total Oracle holdings.

The SLC also concluded that plaintiffs' mismanagement claims against the individual defendants would be difficult to prove at trial. Plaintiffs' primary mismanagement claim is that defendants improperly manipulated revenues. While there was evidence of certain incidents of improper revenue recognition such as backdating contracts, improperly, and removing dates from contracts by lower level employees, there does not appear to be evidence that these practices were prevalent within the company or even known to the individual defendants.

The committee also noted certain circumstances that may have contributed to inadvertent misrecognition of revenue: (1) accounting policies, established when Oracle was a small, non-diversified, non-public corporation, that had become obsolete (due to the growth of the company) by the beginning of the class period; (2) uncertainty in the software industry as to the correct accounting treatment to be given to various software contracts; (3) Oracle's creation of new types of contracts, for which accounting conventions had not yet been developed; and (4) strain on Oracle's financial departments resulting from bunching of contracts at the end of Oracle's fiscal quarters.

As noted above, however, any intentional misrecognition of revenue can probably be attributed only to lower level management and not to the defendant senior managers or directors. Thus, the SLC concluded that even if plaintiffs have a claim against Oracle for mismanagement, they do not have a likelihood of success on their claims against the individual defendants.

With regard to plaintiffs' allegations of fraudulent forecasting of corporate prospects by the individual defendants, the SLC found no evidence of anything but a good faith and reasonable attempt by defendants to project future earnings. The SLC suggested that management's pride in the corporate "track-record" might have played a part in the overly optimistic forecasts issued by Oracle. Nevertheless, there is nothing to indicate that the individual defendants engaged in the kind of intentional or reckless misconduct necessary to constitute fraud.

In light of these difficulties in pursuing the derivative litigation, the SLC recommended settling under the terms of the proposed derivative settlement. The obvious benefit of settlement to the corporation and its shareholders is the termination of all subsequent legal fees and other expenses associated with the derivative litigation. The SLC also believed that settlement will provide Oracle with "the intangible benefit of finality." SLC Rep at 59.

The SLC report considered and rejected the alternative of moving to terminate the derivative action entirely because termination would allow derivative counsel to pursue fees and costs from Oracle in excess of the $750,000 cap for derivative counsel fees established by the proposed derivative settlement. While this seems to be the most questionable conclusion drawn by the SLC (if the derivative claims are so weak, the prospect of plaintiffs recovering attorney fees is slim), it cannot be characterized as an unreasonable one. The costs to Oracle of mounting a successful motion to terminate, and its defense on appeal, probably approach the amount of fees derivative counsel can reasonably anticipate recovering on the present settlement. Even if derivative counsel are awarded some large fraction of $750,000, such a sum could readily be consumed in the fairly extensive proceedings necessary to terminate the derivative litigation. Thus, the SLC's conclusion that settlement under the proposed terms is preferable to moving to terminate falls within the broad latitude allowed by the business judgment rule.

The SLC also concluded that it would be in the best interests of Oracle to grant Arthur Andersen a release as part of the derivative settlement. The committee investigated the possibility of bringing claims against Arthur Andersen for negligence or other wrongdoing due to Arthur Andersen's execution of its annual audits and quarterly reviews of Oracle during the relevant time period. The SLC found insufficient evidence to support any such claims. Moreover, it found that litigation against Arthur Andersen would be costly and would likely cause Arthur Andersen to revoke its offer to contribute $1.75 million to the class action settlement fund. Oracle would then have to pay the additional $1.75 million to settle the class action or lose the benefit of that settlement. For these reasons, the SLC concluded it is not in the best interests of Oracle to pursue claims against Arthur Andersen.

The court has analyzed the SLC's reasoning, reviewed the SLC report and the record, and considered the oral presentations of the parties. The conclusions of the SLC are well reasoned and supported by the record. The committee's recommendations therefore satisfy the business judgment standard.

Furthermore, the court notes that the derivative defendants consist of Oracle's founder and chief executive officer, the company's key executives and an outside director. By any reasonable measure, Oracle has been a highly successful enterprise. Long-term holders of the company's common stock have benefited handsomely from this success. The individual defendants are themselves substantial long-term holders of stock and their possible motivation for manipulating the price of Oracle's stock has never been made clear in this litigation. In any event, it appears that the best interests of Oracle's shareholders are served by allowing these individuals to devote themselves to Oracle's business affairs rather than distracting them with further litigation.

Accordingly, the court hereby APPROVES the derivative settlement. For the reasons stated in Oracle IV, 829 F.Supp. at 1178-83, and because the derivative settlement is approved, the court hereby APPROVES the class settlement.

II

Having approved the final derivative settlement, the court must determine the extent to which derivative counsel are entitled to attorney fees. The derivative complaint alleges state law claims of breach of fiduciary duty and federal claims under various securities laws against defendants. The court has analyzed state and federal law governing derivative counsel fees and has found no substantial differences between Delaware and federal law. The parallel nature of the two applicable bodies of law will become apparent in the discussion below. At this point, however, the court simply notes that it need not determine which law governs the fee determination because Delaware and federal law are in accord on the issue of derivative counsel fees. See Lewis v. Anderson, 692 F.2d 1267, 1270 (9th Cir.1982).

The background rule governing attorney fees in litigation is the "American Rule," which requires a prevailing party to bear its own expenses in the absence of contractual or statutory language to the contrary. See Tandycrafts, Inc. v. Initio Partners, 562 A.2d 1162, 1164 (Del.1989); Reiser v. Del Monte Prop. Co., 605 F.2d 1135, 1137 (9th Cir.1979). Under both Delaware and federal law, a court may grant fees and expenses to derivative counsel when the derivative suit creates a common fund or confers a substantial corporate benefit. Tandycrafts, 562 A.2d at 1164, Mills v. Electric Auto-Lite Co., 396 U.S. 375, 394-96, 90 S.Ct. 616, 626-28, 24 L.Ed.2d 593 (1970). "Typically, successful derivative * * * suits which result in the recovery of money * * * or which result in the imposition of changes in internal operating procedures that are designed to produce such monetary savings in the future, are viewed as fund creating actions." Tandycrafts, 562 A.2d at 1164 (citing CM & M Group, Inc. v. Carroll, 453 A.2d 788, 795 (Del.1982)).

As part of the derivative settlement, derivative counsel agreed to seek no more than $750,000 in fees and costs. In exchange, Oracle agreed not to oppose a fee request of up to $750,000. Nevertheless, the specific fee amount to be granted to derivative counsel is still left to the broad discretion of the court. Tandycrafts, 562 A.2d at 1165; Lewis, 692 F.2d at 1269.

Derivative counsel are entitled to compensation so long as two requirements are satisfied: (1) the derivative suit creates a substantial benefit to the corporation; and (2) the derivative complaint was "meritorious." Chrysler v. Dann, et al., 43 Del.Ch. 252, 223 A.2d 384, 386-87 (1966); Lewis, 692 F.2d at 1270-71. The first requirement is clearly satisfied in this case. At a minimum, it is undisputed that the derivative suit was a causal factor in Arthur Andersen's decision to contribute $1.75 million to the class settlement. Because Oracle would otherwise have had to pay an additional $1.75 million to settle the consolidated class action on the same terms, Arthur Andersen's contribution is a "substantial benefit" to Oracle.

The second requirement is that the complaint must be "meritorious." Chrysler, 223 A.2d at 386-87; Lewis, 692 F.2d at 1270-71. A complaint satisfies this requirement if it would have survived a motion to dismiss on the pleadings. Lewis, 692 F.2d at 1270-71. Although defendants did not move to dismiss the derivative complaint in this action, the court finds such a motion would likely have been unsuccessful. On its face, the complaint alleges facts that sufficiently support claims of insider trading and breach of fiduciary duty. Furthermore, a summary judgment motion by the individual defendants might fail because derivative plaintiffs' claims of mismanagement were supported by evidence of "improper revenue recognition such as backdating contracts, improperly counting contracts, and removing dates from contracts by lower level Oracle employees." SLC Rep at 44. Derivative plaintiffs also had evidence that the director and officer defendants fostered a corporate culture and created monetary incentives that had the effect of encouraging lower-level employees to engage in misconduct. See id at 48-49. Although plaintiffs do not have a likelihood of success on the merits of their mismanagement claims for the reasons noted above in Section I B, the evidence, combined with the inferences of defendants' intent or recklessness that the evidence creates, certainly give derivative plaintiffs a "reasonable hope" of proving intentional or reckless mismanagement by defendants. See Chrysler, 223 A.2d at 256-57 (emphasis added). Thus, the "meritoriousness" requirement for granting derivative counsel fees is satisfied.

Derivative counsel are thus entitled to attorney fees for their services rendered in litigating the derivative suit. The appropriate amount of compensation must be based on the benefit that has been conferred to Oracle by the derivative suit. The court must thus first determine the total benefit that Oracle received from the derivative suit and then determine, from that benefit, a reasonable attorney fee award for derivative counsel.

A

In measuring the benefit to the corporation, it is important to note that the derivative settlement negotiated by derivative counsel did not directly produce any monetary recovery to the corporation. Derivative counsel, however, identify a number of other benefits to Oracle that were allegedly created by the derivative suit. First, derivative counsel argue that the derivative suit led to the $1.75 million that Arthur Andersen agreed to contribute to the class settlement fund.

Second, derivative counsel attribute numerous beneficial changes in Oracle accounting and insider trading policies to the institution of the derivative suit. These "therapeutic changes" supposedly include: (1) changes in Oracle's revenue recognition policies which have given Oracle's financial statements and other pronouncements greater credibility in the financial community; (2) changes in the handling and approval of customer contracts, including the deletion of "four day" signing procedures which created great uncertainty and confusion during the class period; (3) changes in Oracle's insider trading policy including an admonition against speculative transactions in Oracle stock such as dealing in puts, calls, short sales and straddles and a clarification of the definition of "material non-public information." SLC Rep at 63-64.

Even putting aside the obvious difficulty in valuing these policy changes for the purpose of determining the total benefit conferred on the corporation, it is undisputed that there is a serious causation problem in any attempt to attribute the policy changes to the derivative suit. After all, the policy changes could have been and probably were caused by a large number of factors such as the class action, the SEC investigation, negative publicity and other economic forces. Under these circumstances, it is simply impossible for derivative counsel to show, with any degree of credibility, that the policy changes were a result of the derivative suit.

Derivative counsel argue that Tandycrafts shifts the burden of proof to Oracle to show the corrective corporate policy changes were not caused by the derivative suit. In that case, the Supreme Court of Delaware stated that "[o]nce it is determined that action benefiting the corporation chronologically followed the filing of a meritorious suit, the burden is upon the corporation to demonstrate `that the lawsuit did not in any way cause their action.'" Tandycrafts, 562 A.2d at 1165 (quoting Allied Artists Picture Corp. v. Baron, 413 A.2d 876, 880 (Del.1980)). According to derivative counsel, the same rule should be applied in this case to shift the burden of proof to Oracle. Derivative counsel further argue that because Oracle made no substantial showing that the policy changes were not caused by the derivative suit, the court should presume that the policy changes were benefits conferred by the derivative suit. The court disagrees.

Derivative counsel misstate Delaware law when they argue that Oracle has the burden of proving lack of causation. According to the cases cited by Tandycrafts, the burden-shifting rule only applies when the beneficial corporate action completely remedies the wrong complained of and thereby moots the derivative action. See McDonnell Douglas Corp. v. Palley, 310 A.2d 635, 637 (Del. 1973);^[3]Allied Artists Pictures, 413 A.2d at 880.^[4] Thus, in each of these "burden-shifting" cases, the derivative plaintiff's cause of action was made moot by the remedial corporate measures before the burden was shifted to the corporation to show lack of a causal link between the action and the derivative suit.

For example, the derivative plaintiff in McDonnell Douglas sought to have the corporation rescind its sale of 184,127 shares of its own stock to McDonnell Douglas, which owned 56 percent of the common stock of the corporation. Plaintiff complained that the sale was unfairly detrimental to the corporation and unduly beneficial to McDonnell Douglas. Subsequently, the derivative action became moot when plaintiff's corporation was merged into a wholly owned subsidiary of McDonnell Douglas. See Palley v. McDonnell Company, 295 A.2d 762, 763-64 (Del. 1972) (lower court decision). In determining an attorney fee award, the court placed the burden of proof on McDonnell Douglas to show that the merger was not caused by the institution of the derivative suit. See McDonnell Douglas, 310 A.2d at 637.

In Tandycrafts, the derivative plaintiff's complaint was similarly mooted by the corporation's corrective action. The plaintiff in that case sought a preliminary injunction to prevent the corporation from holding its annual meeting, alleging that the issued proxy material regarding certain charter amendments was materially misleading. The corporation subsequently released a supplemental statement rectifying the misleading aspects of the proxy material. The injunction was therefore denied and the charter amendments were later defeated at the annual meeting. The Tandycrafts court shifted the burden of proof to the corporation to show that the derivative litigation did not cause the corrective action. While the court did not explicitly note that the corporation's corrective action mooted plaintiff's complaint, it is clear from the circumstances of the case and from the cases cited that the court was simply applying the "mootness test."

In contrast, Oracle's changes in corporate policy in this case did not moot derivative plaintiffs' complaints of fraud, insider trading and breach of fiduciary duty by the individual defendants. Corporate policy changes would remedy only prospective injury and not past damages Oracle suffered and thus would not moot the derivative claims here. Derivative counsel have pointed to no Delaware cases supporting burden-shifting in the absence of a completely moot derivative plaintiffs' case. Neither have derivative counsel proffered a basis for the court to extend the "mootness test" beyond that which the Delaware courts have heretofore developed.

Moreover, even if the burden of proof to show lack of causation were shifted to Oracle, the burden would be satisfied under the present circumstances. It appears from the record that the changes in Oracle's corporate policies were not caused by the derivative suit. Other causative factors such as the SEC investigation, the class action and public scrutiny were more numerous and considerably more compelling than this rather fecklessly prosecuted derivative action. See, e.g., Chrysler Corp. v. Dann, et. al., 43 Del.Ch. 252, 223 A.2d 384, 389 (1966) (holding that lack of causal connection between alleged corporate benefit and the derivative suit justified denial of fee award).

The Ninth Circuit has analogized an SEC investigation to a grand jury investigation. See Woolley v. United States, 97 F.2d 258, 262 (9th Cir.1938). The negative effects of an SEC investigation on the investigatee are certainly comparable in magnitude to its criminal counterpart. One commentator has described the powerful effect of SEC investigations:

* * * an [SEC] investigation can have a life and impact of its own. It may drag on for years. It may be costly in legal fees, executive and employee time, search and reproduction costs, and other ways. It may be disruptive of business. * * * An investigation may cause customers, employees, creditors, investors or business associates to worry or lose confidence. It will almost always be frightening and aggravating — or worse — to the investigatee. A company under investigation may find it difficult to clear registration statements, proxy statements, or other actions at the SEC. [citation omitted]. Thus an investigation can be much more than a prelude to a possible enforcement action. The investigation can itself be painful and punitive.

5 Alan R. Bromberg & Lewis D. Lowenfels, Securities Fraud & Commodities Fraud, § 13.1(200), p 13:10 (1993). Because an SEC investigation can have such deleterious consequences, it is only predictable that a corporation under investigation would act decisively to resolve the conflict.

The SEC investigation of Oracle began almost concurrently with the filing of the derivative lawsuit. The record shows that the SEC investigated whether certain of the defendants had engaged in insider trading and the "weaknesses in Oracle's system of internal accounting controls." SLC Rep at 34-35. Specifically, the SEC took the depositions of defendants Ellison, Kennedy and Walker. In those depositions, the SEC focussed on whether certain of the individual defendants had engaged in insider trading. Id at 34.

The SEC also investigated Oracle's accounting practices and whether Oracle failed to keep accurate books and records resulting in materially false and misleading financial statements filed with the SEC. It was the SEC's belief that Oracle had engaged in numerous improper accounting and revenue recognition practices which included double billing customers for products and technical support, invoicing customers for work never performed, not issuing credits for customer returns and improperly recognizing contingent or premature revenue. Id at 35.

During the entire investigation, "Oracle actively worked with the SEC to identify and resolve issues." Id at 35. Together, they worked toward a settlement of the SEC's alleged claims against Oracle. On June 1, 1992, Oracle implemented its new insider trading and revenue recognition policies. On September 24, 1993, the SEC filed a complaint alleging that Oracle had engaged in misleading accounting and revenue recognition practices. Concurrently, Oracle filed a consent to the entry of final judgment enjoining it from engaging in future violations as alleged in the SEC Complaint and agreeing to pay a civil penalty in the amount of $100,000. Id at 37.

Oracle's pattern of acquiescence toward the SEC during the investigation is clear, and indeed understandable in light of the aforementioned negative effects of SEC investigations. Oracle's remedial policy changes appear simply to be one step in the process by which Oracle hoped to placate the SEC. Needless to say, the SEC's in-depth investigation into insider trading by Oracle officials would, without more, easily have motivated Oracle to change its internal safeguards against insider trading. Moreover, based on the SEC's oft-stated position that Oracle's revenue recognition practices were not in compliance with generally accepted accounting practices or federal law, Oracle was essentially compelled to change its flawed accounting and revenue recognition policies in order to prevent the SEC from prosecuting Oracle for future alleged violations of securities laws. Thus, most, if not all, of Oracle's accounting and insider trading policy changes appear to have resulted from SEC pressure.

That leaves the Arthur Andersen contribution of $1.75 million as the sole identifiable benefit that may have been conferred by the derivative action. Even this is at least somewhat doubtful. Arthur Andersen may very well have made the same contribution to the class settlement in the absence of the derivative action. That sum is a relatively small amount for a major accounting firm to pay in settlement of significant litigation, especially a securities class action, the defense of which alone would probably have generated legal fees and expenses close to the settlement amount. Still, because the derivative action no doubt played some role in precipitating Arthur Andersen's contribution and because it is possible for derivative counsel to establish a causal nexus, it is fair to give derivative counsel the benefit of the doubt and ascribe the Arthur Andersen contribution to the derivative litigation.

B

A remaining issue is the amount that should be granted to derivative counsel as attorney fees and costs given the conclusion that the derivative litigation conferred a $1.75 million benefit upon Oracle.

1

There are two general approaches used to determine attorney fee awards in derivative suits. First, in a case such as this one, where the litigation results in a "common fund" that benefits the general class of shareholders, courts may allocate a percentage of the total fund as the fee award. See, e.g., Sugarland Industries, Inc. v. United States National Bank of Galveston, 420 A.2d 142, 149 (Del.1980); Six Mexican Workers v. Arizona Citrus Growers, 904 F.2d 1301, 1311 (9th Cir.1990).^[5] The second is the so-called "lodestar" approach in which the fee is determined based on the time and materials spent by derivative counsel. See Loring v. City of Scottsdale, Ariz., 721 F.2d 274, 275-76 (9th Cir.1983) (citing Lindy Bros. Builders, Inc. of Philadelphia v. American Radiator & Standard Sanitary Corp., 487 F.2d 161 (3d Cir.1973)). Either of these methods, depending on the circumstances, can be used to determine a reasonable fee. See Six Mexican Workers, 904 F.2d at 1311.

Under Delaware and federal law, the court has broad discretion to choose the appropriate method for determining the fee award. Sugarland Industries, 420 A.2d at 149; Six Mexican Workers, 904 F.2d at 1311. In Delaware, the following factors are pertinent to the derivative counsel fee determination:

the `results achieved' by counsel plus * * * the amount of time and effort applied to a case by counsel for plaintiff, the relative complexities of the litigation, the skills applied to their resolution by counsel, as well as any contingency factor and the standing and ability of petitioning counsel * * *.

Sugarland Industries, 420 A.2d at 149; Id. In Sugarland Industries, the Delaware Supreme Court found a "20%-of-benefit factor" reasonable. Id. at 151.

The Ninth Circuit takes the same general approach as the Delaware Supreme Court outlined in Sugarland, but has established 25 percent of the fund as the "benchmark" fee award in common fund cases. See Six Mexican Workers, 904 F.2d at 1311. "The benchmark percentage should be adjusted, or replaced by a lodestar calculation, when special circumstances indicate that the percentage of recovery would be either too small or too large in light of the hours devoted to the case or other relevant factors." Id. (emphasis added). The factors identified in Sugarland Industries, 420 A.2d at 149, and quoted in the preceding paragraph, likewise qualify as "relevant factors" under the Ninth Circuit approach.

While the common fund benchmark approach is not binding on this fee application (which is governed by Delaware and federal law applicable to derivative litigation) it is not inconsistent with Delaware law or other Ninth Circuit cases on this issue.^[6] Because this case is essentially a common fund case, see supra note 5, and the court did not set derivative counsel's compensation, unlike that of class counsel, at the outset of the case, the court must employ an ex post assessment of a reasonable fee for derivative counsel. Despite its shortcomings, see Oracle I, 131 F.R.D. at 695-97, the "benchmark" percentage fee approach still enjoys judicial acceptance and its alternative, the lodestar approach, carries its own numerous shortcomings. Id. at 689-93. Furthermore, derivative counsel implicitly rejected the lodestar approach themselves when they agreed to cap their fees at $750,000.

Consistent with derivative counsel's implicit rejection of it, the court declines to use a lodestar approach to calculate derivative counsel's fees. See, e.g., Sugarland Industries, 420 A.2d at 150 (upholding lower court's rejection of "lodestar" in favor of the percentage of the fund). Although the court will use the percentage of the fund approach to determine a reasonable fee award for derivative counsel, this choice does not, as will presently be explained, eliminate derivative counsel's effort as a factor to be considered.

2

Derivative counsel object to using the 25 percent benchmark approach because it allegedly devalues the time and effort they have devoted to the case. See Deriv Csl Decl in Supp of Fees filed March 22, 1993, at ¶ 32, and Ex E, F, G, H. Yet the Ninth Circuit has made clear that courts may adjust a percentage fee award upward or downward to take into account any equities created by the attorney time and effort devoted to the case. See Six Mexican Workers, 904 F.2d at 1311. In reality, therefore, derivative counsel's objection is less methodological and more quantitative. Whatever its characterization, derivative counsel's argument is not wholly insubstantial, however.^[7]

No matter what approach — lodestar or percentage of the fund — is employed, the fee awarded must be a reasonable one. In re Washington Public Power Supply, 19 F.3d 1291, 1294 (9th Cir.1994). "The object in awarding a reasonable attorney's fee * * * is to give the lawyer what he would have gotten in the way of a fee in an arm's length negotiation, had one been feasible. In other words the object is to simulate the market where a direct market determination is infeasible." Matter of Continental Illinois Securities Litig., 962 F.2d 566, 572 (7th Cir.1992). See Swedish Hosp. Corp. v. Shalala, 1 F.3d 1261, 1269 (D.C.Cir.1993).^[8]

In the case of ex post determinations, "reasonableness" is based upon results achieved in light of the efforts of counsel. In an ex ante determination, such as that used to set class counsel's fees here, the competitiveness of the process from which the fee proposals emerged establishes the reasonableness of counsel fees. Because a competitive process for setting derivative counsel fees was not employed at the outset of the derivative litigation, an ex post analysis of results and counsel efforts is necessary.

Under both Delaware and Ninth Circuit law, the step which follows determination of the benchmark percentage fee is to look at the factors identified above to determine whether a departure from that benchmark is justified. Sugarland Industries, 420 A.2d at 149. Six Mexican Workers, 904 F.2d at 1311. As noted above, actual time and resources spent by counsel are relevant to this determination. Derivative counsel's work substantially in excess of the value represented by the benchmark percentage is a factor which supports lifting the percentage award above the benchmark.

In this case, there are also other factors that militate toward a higher fee award than the 25 percent benchmark would obtain: (1) the complexity of this case; (2) the experience, reputation, and ability of derivative counsel; (3) the fact that the potential of recovering a large monetary award from the individual defendants was limited by indemnity agreements; and (4) the high contingent risk that derivative counsel would not succeed on the merits of this litigation. See Sugarland Industries, 420 A.2d at 149. Cf Washington Public Power Supply, 19 F.3d at 1300 n. 10.

Still, these factors do not warrant exceeding the upper limits of the benchmark range. In Six Mexican Workers, 904 F.2d at 1311, the Ninth Circuit cited 3 Newberg on Class Actions, § 14.03 for the proposition that 20 to 30 percent of the fund is the usual range of common fund fee awards in securities and antitrust suits. As noted above, the beginning point for analysis is the 25 percent of the fund "benchmark." Here, that percentage would produce an award of $437,500 (25 percent of $1.75 million).^[9] A departure from the 25 percent benchmark to 30 percent to account for the above factors is appropriate in this case, but an award above 30 percent is plainly not justified.

An award figure exceeding 30 percent of the recovery would produce extraordinary compensation of derivative counsel. Based on the 30 percent figure, derivative counsel would receive $525,000 (30 percent of $1.75 million) in attorney fees. Derivative counsel's substantiation of expenses suggests that derivative counsel are also entitled to reimbursement of expenses in the amount of $69,384.76. See Deriv Csl Decl in Supp of Fees filed March 22, 1993, at ¶ 32, and Ex E, F, G, H. Derivative counsel's total fee and expense award would thus amount to 34 percent of the total fund generated by the derivative suit. Such an award is not merely "reasonable," but borders on the superfluous. A recent empirical study of attorney fees in securities litigation, Stuart J. Logan & Beverly C. Moore, Jr., Attorney Fee Awards In Common Fund Securities & Antitrust Class Actions, 13 Class Action Reports 249, 546 (1990), analyzed 334 common fund securities cases and found that in the 48 cases that generated a common fund between $1 million and $2 million, the attorneys received 28.3 percent of the fund as an average fee and expense award. Id. An award of 34 percent of the recovery here for fees and costs represents a 20 percent premium over the average award.

The court does not doubt that derivative counsel have expended time and effort on the case that, at their so-called "normal" billing rates, would generate a fee award above 34 percent of the recovery. But it is unusual for lawyers actually to collect the full amount of their "normal" hourly charges (indeed, these charges might more accurately be termed "abnormal").

The reasonableness, if not generosity, of a $525,000 fee award stems from several circumstances. Because reasonableness, even when determined ex post, must be based on an attempt to replicate what fee bargain would have been struck ex ante, Continental Illinois Securities Litig., 962 F.2d at 572, it is appropriate to observe that most percentage fees agreed to by lawyers and clients for contingent percentage fee litigation in the United States fall in the range between one-quarter and one-third of the recovery. Cf Six Mexican Workers, 904 F.2d at 1311. It is rare that clients agree to share more than one-third of a litigation recovery with their lawyers.

Additionally, the American Rule on attorney fees has justly been criticized for creating undue incentives to litigate by insulating the attorneys and parties who bring unmeritorious lawsuits or assert unmeritorious defenses from claims for recovery of the attorney fees and costs which such tactics inflict on others. Accordingly, courts and attorney disciplinary bodies have been fairly rigorous in enforcing upper limits on percentage contingent fees. Attorneys in contingent fee litigation are forbidden from keeping all or even the greater part of the recovery, despite the deterrent function of litigation being as well (maybe even better)^[10] served if the attorney obtains the whole recovery as only part of it.

The constraints against "unreasonable" or "unconscionable" fees are deeply ingrained features of bar rules governing professional responsibility. See, e.g., ABA Model Rules of Prof Conduct, Rule 1.5(a); Cal Rules of Prof Conduct, Rule 4-200(A). To be sure, ethical prohibitions against "unreasonable" percentage fees serve the self-interest of the legal profession. The client, after all, is unlikely to be motivated to help the lawyer by testifying and producing evidence unless the client retains a stake in the outcome of the case sufficient to animate such assistance. Hence, limitations on percentage attorney fees are necessary to the viability of the contingent fee system.

From this perspective, limitations on allowable percentage fees might therefore seem unnecessary in representative litigation because clients play no important role in that type of litigation. But this would ignore another purpose served by these ethical limitations that clearly applies to representative litigation: discouragement of litigation asserting insubstantial claims. If there were no limitation on percentage fees, longshot or trifling claims would be encouraged and crowd out of the judicial queue cases which have a reasonable probability of success or importance. The perception that this already occurs explains much of the commentary critical of class actions. See Dave Barry, Lawyers Get Their Teeth into Some Real Class Action, Chicago Tribune, Sunday Magazine, Dec 5, 1993, at 33 ("It cannot be easy, taking a case wherein it appears, to the naked untrained layperson eye, that nobody has suffered any observable harm, and, using legal skills, turning it into a financial transaction that involves thousands of people and a million dollars! Plus coupons!").

An upper limit to the percentage of recovery that contingent fee counsel can be compensated for their risk in both ordinary client-driven litigation and in lawyer-driven litigation such as class and derivative actions serves this purpose. Because potential aggregate recoveries can be so enormous and hence the dangers of extortionate settlements so grave in class and derivative litigation, the need for upper limits on percentage compensation of class or derivative counsel is at least as compelling as in the case of ordinary litigation. Unbounded percentage fee awards would unduly encourage lawyers to bring class or derivative suits too risky to countenance in a sensible judicial system. Cf City of Burlington v. Dague, ___ U.S. ___, ___, 112 S.Ct. 2638, 2641, 120 L.Ed.2d 449, 457 (1992) (noting, in the fee-shifting context, that "enhancement for the contingency risk posed by each case would encourage meritorious claims to be brought, but only at the social cost of indiscriminately encouraging nonmeritorious claims to be brought as well").

This means that there cannot be a one-to-one correspondence between fully risk-adjusted fees and what courts deem to be reasonable percentage fees. The risks required to increase (lower) the percentage above (below) the benchmark should rise (fall) faster than the change in the percentage fee. Otherwise, encouragement of longshot and trivial cases would be too great or attorneys would be shortchanged in their efforts. Although the 30 percent of recovery figure may not be an absolute ceiling, it is plain that only compelling circumstances warrant an award above that percentage. While derivative counsel have pointed to some factors which augur for a premium to the 25 percent "benchmark," the factors identified by derivative counsel are hardly compelling enough to warrant breach of the 30 percent upper range.

Accordingly, derivative counsel are hereby AWARDED $525,000 in attorney fees and $69,384.76 in expenses.

III

Because the class action settlement has been approved, the time has come for the court to revisit the subject of class counsel compensation. In marked contrast to derivative counsel fees, which had to be calculated ex post using various judicial doctrines of "reasonableness," class counsel fees and costs in this case were determined ex ante as a result of competitive bidding among the various law firms that sought to represent the class.

On August 3, 1990, the court solicited competitive bids from numerous law firms seeking to represent this class as class counsel to prosecute claims against all defendants except Arthur Andersen ("Oracle defendants"). See Oracle I, 131 F.R.D. 688 (N.D.Cal.1990). Four firms submitted blind bids in response to the court's request. Oracle II, 132 F.R.D. 538, 539 (N.D.Cal.1990). The bidders were highly competent law firms, each with extensive experience in prosecuting securities class actions.

After reviewing the bids, the court chose the law firm of Lowey, Dannenberg, Bemporad, Brachtl & Selinger, P.C., to serve as class counsel. Pursuant to its bid (the "Oracle bid"), the Lowey firm represented the class against the Oracle defendants under a basic fee schedule with an early settlement discount, and a $325,000 cap on litigation expenses to be charged to the class:

Time for Resolution (months)
Recovery             0-12        13 or more
(Lowey # 1)     (Lowey # 2)
____________________________________________
Up to $1M           24%             30%
$1M-$5M             20%             25%
$5M-$15M            16%             20%
$15M or more        12%             15%

Id. at 541.

By early 1991, it became clear that Arthur Andersen ("Andersen") would be named an additional defendant in these consolidated actions. Accordingly, the court again solicited competitive bids from law firms to determine class counsel to prosecute class claims against that defendant. Order dated June 15, 1991. Three highly qualified law firms submitted bids. After comparing the bids, the court issued an order on July 21, 1991, designating the Lowey firm as class counsel for the claims against Andersen. Order dated July 21, 1991.

The Lowey bid for the Andersen claims established the same fee scale that Lowey had proposed for its representation of the class claims against the Oracle defendants, but without an early settlement discount (the "Andersen bid"):

Recovery          Percentage
the first $1M          30%
the next $4M           25%
the next $10M          20%
excess of $15M         15%

Under the terms of Lowey's bid for the claims against Andersen, any recovery from Andersen is aggregated with other recoveries for purposes of determining class counsel's compensation. Additional litigation expenses attributable to the Andersen claims were capped at $500,000.

This section addresses three issues relating to class counsel fees. First, pursuant to the July 21, 1991, order, the court will discuss its rationale for selecting the Lowey firm as counsel to prosecute the class claims against Andersen. Second, the court will resolve class member Teeters' motion for reconsideration of the court's order accepting a percentage of the fee approach to determine class counsel fees. Third, the court will calculate class counsel's fees and costs award and review its appropriateness.

A

As noted above, the court received three bids from firms which sought to represent the class against Andersen. The firms that submitted bids were: (1) the Lowey firm; (2) Stamell, Tabacco & Schager; and (3) David B. Gold, PLC. All three of these firms were qualified to represent the class against Andersen. The court conducted an in-depth comparison of the bids and found that the Lowey and Stamell bids were significantly lower in price at almost all time and recovery contingencies than the Gold firm's bid. An analysis also revealed that the Stamell bid was somewhat superior in price to the Lowey bid in the event the case were resolved prior to jury selection. After jury selection, however, the Lowey bid became the more attractive bid based on price alone.

Because the Lowey and Stamell bids did not differ significantly in terms of price, the court turned to other factors to choose a bid. As the court noted in prior orders, the selection of class counsel through the bidding process need not, and should not be, a function only of price. See Oracle III, 136 F.R.D. 639, 648-50 (N.D.Cal.1991).

The bidders failed to provide qualitative information sufficiently tangible to differentiate themselves. See Oracle III, 136 F.R.D. at 648-49. Nevertheless, two factors convinced the court to choose the Lowey bid. First, it appeared more efficient to have one firm prosecute the entire case on behalf of the class. At the time, the Lowey firm had already been chosen to represent the class on claims against the Oracle defendants and had spent considerable amounts of time and resources familiarizing itself with the case. Undoubtedly, the Lowey firm's experience in prosecuting the claims against the other defendants had given that firm knowledge and background directly applicable to the claims against Andersen. Second, the Lowey firm already had a "proven track record" in this case by the time the court had to choose counsel to prosecute the class claims against Andersen. The Lowey firm's high calibre representation of the class against the Oracle defendants served as tangible evidence of its ability to represent the class against Andersen. These factors warranted selection of the Lowey firm to serve as counsel on the class claims against Andersen even though its bid may have been somewhat higher than that submitted by the Stamell firm.

B

Class member David Teeters objects to the court's use of a percentage of the common fund contingent fee approach to determine class counsel fees. Teeters claims that he has a right to make his objections orally in a rescheduled settlement hearing, but cites no authority for that proposition. To the contrary, Local Rule 220-1 authorizes the court to decide a motion with or without oral argument. Teeters has had numerous opportunities to submit his objections in writing, see, e.g., Order dated Dec. 21, 1993 ("If any party or objector wishes to comment on the issue of class counsel fees * * *, he or she must submit the comments, in writing, to the court by January 11, 1994."), and has in fact made a complete presentation of his objections in his written submissions. See, e.g., Teeters' "Notice of Intention to Appear and Object to * * * Award of Attorneys' Fees" dated March 31, 1993. Accordingly, Teeters' motion is deemed submitted on the papers pursuant to Local Rule 220-1.

Teeters' objections boil down to his contention that a percentage fee method of awarding class counsel fees from a common fund is inappropriate because it necessarily incorporates a risk multiplier. City of Burlington v. Dague, ___ U.S. ___, ___, 112 S.Ct. 2638, 2643, 120 L.Ed.2d 449, 459 (1992). According to Teeters, Dague and other cases mandate that the court use a lodestar method without a risk multiplier method to calculate class counsel fees. The court is unpersuaded by this argument.

The propriety of using a percentage of the fund contingent fee approach in this case was addressed at the outset of the litigation. Oracle I, 131 F.R.D. at 694-95. At that time, the court weighed the relative advantages and disadvantages of the lodestar and percentage of the fund approaches in the common fund context and found percentage of the fund fee compensation "vastly superior." Id. at 694.

Specifically, the court noted that a percentage of the fund contingent fee is superior as a monitoring device because percentage of the fund better "`align[s] the interest of lawyer and client. The lawyer gains only to the extent his client gains.'" Id. (quoting Kirchoff v. Flynn, 786 F.2d 320, 325 (7th Cir. 1986)). Because the lawyer has incentives similar to his client, there is less need to monitor the lawyer's performance. Id.

The monitoring function of percentage fees is particularly appropriate for class actions. Class members typically have such small individual stakes in the litigation that they have little (or, more accurately, no) incentive to monitor the litigation, much less assess the quality of class counsel's performance. Neither are courts in the position to monitor class counsel's performance during the litigation as that would entail disclosure of the weaknesses and shortcomings of the plaintiffs' case and other highly confidential information. After a settlement or trial, courts lack any real incentive to evaluate class counsel's performance, by which time it is too late for such monitoring to do any good, anyway. A percentage of the common fund thus represents a reasonable substitute for the lack of client and judicial monitoring in class actions. For these reasons, the court decided that class counsel in this case should be compensated on a percentage of the fund basis.

Dague and the other cases cited by Teeters do not alter this analysis. In Dague, the Supreme Court held that a court, in determining an award of attorney fees under a federal fee-shifting statute, could not enhance the fee award above the lodestar amount to compensate for the risk of nonpayment. ___ U.S. ___, 112 S.Ct. 2638, 120 L.Ed.2d 449 (1992). The Court articulated three "reasons for concluding that no contingency enhancement whatever is compatible with the fee-shifting statutes at issue." Id., ___ U.S. at ___, 112 S.Ct. at 458. "First, just as the statutory language limiting fees to prevailing * * * parties bars a prevailing plaintiff from recovering fees relating to claims on which he lost, [cite omitted], so should it bar a prevailing plaintiff from recovering for the risk of loss." Id. Second, in fee-shifting cases, the courts have "generally turned away from the [percentage of recovery] contingent-fee model * * * to the lodestar model." Id., ___ U.S. at ___, 112 S.Ct. at 458-59.^[11] Third, the policy favoring "administrability" militates against adoption of contingency enhancement which "would make the setting of fees more complex and arbitrary, hence more unpredictable, and hence more litigable." Id., ___ U.S. at ___, 112 S.Ct. at 459.

Dague is inapposite to the case at hand because Dague only applies in the fee-shifting context, and not the common fund context. Swedish Hosp. Corp. v. Shalala, 1 F.3d 1261, 1268 (D.C.Cir.1993). In Swedish Hosp. Corp., the District of Columbia Circuit carefully considered and rejected the proposition that "[Dague] * * * mandate[s] an unenhanced lodestar approach in common fund cases." That court held that common fund cases are sufficiently different from fee-shifting cases so that the Dague rationale has no applicability in the former context. Id. The court agrees with this conclusion.

Judge Posner articulated the critical difference between common fund and fee-shifting suits best in Matter of Continental Illinois Securities Litig., 962 F.2d 566, 573 (7th Cir.1992):

The problem with risk multipliers in [fee shifting] cases is that the lawyer has two sources of fees: the statutory fee award, and the contingent-fee contract that he negotiates with his client. The latter takes care of the risk of loss and thus makes the inclusion of a risk multiple in the statutory award redundant, assuming as the cases do that the purpose of fee-shifting is to assure competent representation for plaintiffs rather than to make the cost of litigation to the successful plaintiff zero. [cites omitted]. In a class action suit, however, there is no contract between lawyer and client — no source other than the judicial award of fees to which the lawyer can look for compensation for risk of loss. The situation is thus parallel to that in an ordinary damages case in which the prevailing plaintiff has no entitlement to a judicial award of fees. The difference of course is that instead of the market's dictating the terms on which the lawyer is retained, the judge must do so. But he must try as best he can to simulate the market * * *.

In other words, risk enhancements in the fee-shifting context are inappropriate and redundant because the attorney can seek compensation for the risk of non-payment from his client. In a common fund class action, however, counsel has only one way to compensate for the risks of litigation — by receiving it in the judicial award of fees.

Because of the fundamental difference between fee shifting and common fund cases, the Dague rationale cannot be applied to this case. Mandating the use of the unenhanced lodestar method to calculate class counsel's fees in this common fund case would close off class counsel's lone source of compensation for the risk of non-payment. Such a result could not have been intended by the Supreme Court. Dague, after all, did not foreclose counsel even in a fee-shifting context from seeking compensation for the risks of litigation in a private contract with their client. There is no reason to do so in common fund litigation.

Furthermore, Teeters' argument that class counsel are not entitled to be compensated for the risks of litigation runs counter to current Ninth Circuit case law. In the case of Torrisi v. Tucson Elec. Power Co., 8 F.3d 1370, 1376-77 (9th Cir.1993), the Ninth Circuit ratified the use of a percentage of the fund approach to compensate class counsel. That court held that class counsel were entitled to more than the unenhanced lodestar amount because class counsel had the case on a "double contingency; first, they must prevail on the class claims, and then they must find some way to collect what they win." Id.

Indeed, the Ninth Circuit recently held that a district court's refusal to enhance class counsel's lodestar fee award in a common fund case by using a risk multiplier was an abuse of discretion. In re Washington Public Power Supply, 19 F.3d 1291 (9th Cir. 1994). The court in that case rejected class plaintiffs' argument that Dague should be extended to prohibit contingency enhancements in common fund cases. Id., 19 F.3d at 1299-1300. The Washington Public Power Supply court held that the Dague rationale did not apply in the common fund context because "[u]nlike statutory fee-shifting cases, where the winner's attorneys' fees are paid by the losing party, attorneys' fees in common fund cases are not paid by the losing defendant, but by members of the plaintiff class * * *." Id., 19 F.3d at 1300. The court reasoned that while it may be inequitable to burden a defendant with a contingency enhancement "that has the effect of compensating the winning attorneys for time gambled away in the contingency cases they lose," it is not unfair to allow a contingency enhancement in common fund cases because those who benefit from the creation of the fund [i.e., the class plaintiffs] should "share the wealth with the lawyers whose skill and effort helped create it." Id., 19 F.3d at 1300.

For the above reasons, Teeters' reading of Dague is untenable. The Ninth Circuit, along with the Seventh Circuit and District of Columbia Circuit, all authorize the use of a percentage of the fund approach to determine class counsel fees in common fund cases. Accordingly, the court hereby reaffirms its original decision to grant attorney fees based on a percentage of the fund calculation and DENIES Teeters' motion for reconsideration.

C

The actual calculation of class counsel fees is fairly straightforward due to the existence of class counsel's pre-determined fee arrangements. As these fee schedules were the product of competitive bidding processes that adequately simulated the market for legal services, class counsel's fee applications are reasonable ex ante. There is no further need to ascertain whether the class counsel fees, to be determined now under the preset fee schedules, are "reasonable" under the standards used in the Ninth Circuit to analyze fees that are judicially awarded ex post. See Oracle II, 132 F.R.D. at 548.

Because, however, use of competitive bidding to select class counsel in this litigation was fairly novel, it is interesting to evaluate the end results of that process. Bids were solicited from a large number of leading plaintiffs' class action law firms that sought to represent the class in this action. See, e.g., Oracle I, 131 F.R.D. at 690. The law firms which submitted bids were highly competent and experienced in securities litigation. Most of the bids allowed for varying percentages of recovery depending on one or more recovery, time or event contingencies similar to those produced by exhaustive judicial review of ex post fee applications. See, e.g., Oracle II, 132 F.R.D. at 539-541. The sophistication of the bids clearly reflected each bidder's experience in assessing the costs and risks associated with securities litigation. The bids were analyzed and compared using multiple criteria before the court made its choice of class counsel. The Lowey bids ultimately selected by the court offered class counsel performance incentives while simultaneously protecting against lawyer windfalls. For the reasons stated in this order and in Oracle II, the court is convinced that competitive bidding adequately simulated the market for legal services and has resulted in "reasonable" fees.

Under the Lowey bids, class counsel fees of $4,800,000 are payable from the $25,000,000 settlement recovery, calculated as follows:

Amount of Recovery               Percentage             Fee
On the First $1 million               30%              $   300,000
On the Next $4 million                25%              $ 1,000,000
On the Next $10 million               20%              $ 2,000,000
On the excess above $15 million
(i.e. $10 million)                  15%              $ 1,500,000
___________
TOTAL:   $4,800,000

Because the case did not settle within twelve months, the early settlement discount of 20 percent was not achieved. Additionally, because the claims against the Oracle defendants and Andersen were settled on a global basis, fees are calculated on the gross recovery to the class.

Class counsel's entitlement to reimbursement of expenses from the fund is limited, under the terms of its bids, to an aggregate amount of $825,000. The Lowey bids established caps of $325,000 on expenses incurred in prosecuting the class claims against the Oracle defendants and $500,000 on expenses incurred in pursuing the class claims against Andersen.

The court has reviewed class counsel's substantiation of their litigation expenses and finds that class counsel incurred $320,065.95 in expenses on the Oracle claims and $472,342.43 in expenses on the Andersen claims (these expenses equal $791,408.38 in the aggregate). See Class Counsel's Mem in Supp of Appl for Fees and Exp at 5-6. Additionally, class counsel recorded $188,176.77 in notice expenses. These costs were incurred in the course of giving notice of the settlement to class members and reimbursing brokers. Id. at 6-7; Notice of Amend of Prior Exp Appl at ¶ 2, and Attached Suppl Affdvts. Of this amount, only $33,591.62 can be reimbursed under Lowey's expense caps ($825,000 - $791,408.38). See Oracle IV, 829 F.Supp. at 1178 n. 3. Finally, class counsel seek reimbursement of $174,176 in expenses incurred as of December 28, 1993, for processing 24,446 proofs of claim. Suppl Affdvt of Dannenberg re Expenses at ¶ 7, Ex 2.^[12] This amount includes the costs of reviewing proofs of claim and sending out deficiency letters. The court has already determined that costs of administering the settlement fall within class counsel's expense caps. See Oracle IV, 829 F.Supp. at 1178 n. 3. Class counsel's request that the court reconsider that determination rests largely on the tax consequences of the characterization of those expenses. See Ltr of Neil L. Selinger, May 10, 1994. As with the costs of giving notice to the class, however, costs of administering and processing claims to the fund are no less related to the conduct of the litigation and hence expenses covered under the expense caps. Because the upper limit of each expense cap has been reached, the balance of the costs of notifying the class and administering the fund must be absorbed by counsel. Accordingly, class counsel's request for reconsideration is hereby DENIED. Class counsel are therefore entitled to reimbursement of only $825,000 of their total expenses.

Class counsel also claim a share of the interest that has accrued on the settlement fund during 1993 and early 1994. According to class counsel's interpretation of its predetermined fee arrangements with the court, class counsel are entitled to 15 percent of all of the interest that has accrued on the settlement fund. Class counsel calculate their award based upon the theory that interest increases the common fund beyond its original amount of $25 million. Because class counsel are entitled to 15 percent of any recovery in excess of $15 million, they claim entitlement to 15 percent of each additional dollar accruing to the fund due to interest. Class counsel's interpretation of its fee arrangements seems flawed and, curiously enough, to their disadvantage.

Class counsel's fee agreements with the court make no explicit provision for interest. In contingent attorney fee arrangements, however, interest on the fund generated by the litigation is typically apportioned pro-rata between counsel and their clients based on their respective percentage ownership of the recovery. As contingent fee lawyers, class counsel are in every sense part-owners of the litigation claims in suit and entitled to the interest accruing on that portion of the claims in which they hold an ownership or equity interest.

In the absence of any explicit contrary arrangement, the standard approach to interest should apply. In this case, class counsel will receive an award of interest to the extent that interest arises directly out of their portion of the settlement (i.e., class counsel's fees and expenses award). Class counsel are entitled to 22.5 percent of the total settlement fund excluding any interest^[13] and are, correspondingly, entitled to the interest that accrued on class counsel's 22.5 percent portion of each settlement installment from the time it was paid by the defendants into the settlement fund.

D

As stated above, competitive bidding eliminates the need to make an after-the-fact determination that class counsel's award of fees and costs is reasonable. Nevertheless, the court notes that ex ante competitive bidding was effective in this case in producing "reasonable" fees from an ex post perspective as well.

By inviting competitive bidding, the court was able to "shop around" and thereby received a fee schedule that represented substantial savings to the class. The court notes the amount of fees (and the percentages of the recovery) that would have been payable under the four bids submitted pursuant to Oracle I, 131 F.R.D. 688 (N.D.Cal.1990) based upon a $25 million settlement at this stage in the proceedings:

Lowey[14]    Abbey[15]   Berger[16]   Gold[17]
$4.8 M          $5.625 M       $6.0 M          $6.5 M
(19.2%)         (22.5%)        (24%)           (26%)

See Oracle II, 132 F.R.D. at 539-42 (describing the details of each submitted bid).

Class counsel collectively expended $3,810,135 worth of unenhanced lodestar time in prosecuting this class action. See Supp Affdvt of Dannenberg re Expenses at ¶ 4. Under the lodestar approach, courts usually scrutinize the hourly charges submitted by class counsel and reduce them significantly. Even in the absence of such scrutiny here, class counsel's fee award of $4,800,000 is reasonable in light of this unenhanced lodestar figure. Because the class claims were not particularly strong on the merits and there was a material risk of non-recovery in this case, the lodestar amount would have to be enhanced with a risk multiplier. See In re Washington Public Power Supply, 19 F.3d 1291, 1294 (9th Cir.1994). A risk multiplier of approximately 1.25 would bring the unreduced lodestar figure up to the fee amount determined by the bidding procedures here. While this is a somewhat lower multiplier than is typically awarded in cases of this type, the total hourly charges have not been subjected to the judicial scrutiny which usually results in a substantial reduction. Such a reduction would correspondingly increase the multiplier and only a fairly modest reduction in hourly charges would bring the multiplier within the range of multipliers courts have typically found appropriate for cases of this risk-magnitude.^[18]

Finally, some lawyers in the litigation questioned whether selection of class counsel through competitive bidding would produce class representation of lesser quality than the traditional approach of delegating class representation to a steering committee of the plaintiff lawyers. See Oracle III, 136 F.R.D. at 647-50. Those questions appear plainly to have been put to rest, at least as far as this litigation is concerned.

By the standards which have typically been used to measure such things, the performance of class counsel in this litigation compares favorably to that of class counsel in other major securities class actions. First, class counsel's recovery of $25 million is "a relatively good class settlement" when measured as a percentage of claimed damages (as distinguished from the highly suspect claims of "trading losses"). Beverly C. Moore, Jr. & Stuart J. Logan, Current Decisions Survey, 16 Class Action Reports 389, 409 (1993). Plaintiffs' expert estimated actual damages during the relevant period^[19] to be approximately $102 million. Hammerslough Decl at ¶ 63. The class recovery represents a respectable 24.5 percent of that estimate. Compare Janet Cooper Alexander, Do the Merits Matter?: A Study of Settlements in Securities Class Actions, 43 Stan L Rev 497, 517 (1991). More importantly, defendants' expert estimated actual damages to be as low as $32.6 million. Ross Decl in Supp of Settlmnt at 2-6. Under that estimate, the $25 million settlement represents a remarkable 77 percent of damages. Second, no officers and directors' liability coverage was available to fund any part of the recovery in this case. Usually, such insurance funds a substantial portion of the recovery. Third, class counsel's efforts were complicated by the necessity to fight a reargued action against a disappointed bidder for class counsel, see Oracle III, 136 F.R.D. 639 (N.D.Cal. 1991), fend off a class action gadfly after the settlement, see supra pp. 1454-1456, and, most importantly, work through undoubtedly more intricate settlement negotiations and endure a long delay in settlement approval due to complications introduced by the companion derivative litigation which was maintained by competing plaintiff lawyers, see Oracle IV, 829 F.Supp. 1176 (N.D.Cal.1993). Class counsel's able, persistent and patient performance in achieving the present recovery has not gone unnoticed.

In sum, class counsel are hereby AWARDED attorney fees in the amount of $4,800,000 and shall have their expenses REIMBURSED from the settlement fund in the amount of $825,000. Class counsel are also AWARDED interest on their fees and expenses in an amount to be calculated in accordance with this order.

Class counsel are directed to prepare, serve and file a proposed form of judgment, objections to which must be served and filed within ten days.

Neil L. Selinger's letter to the court dated May 10, 1994, and the exhibits attached thereto, are hereby ordered filed in the record.

IT IS SO ORDERED.

NOTES

[1] The history of this case is well-documented:

In re Oracle Securities Litig., 131 F.R.D. 688 (N.D.Cal.1990) ("Oracle I") (establishing competitive bidding process to determine class counsel);

In re Oracle Securities Litig., 132 F.R.D. 538 (N.D.Cal.1990) ("Oracle II") (selecting class counsel from bids submitted pursuant to Oracle I);

In re Oracle Securities Litig., 136 F.R.D. 639 (N.D.Cal.1990) ("Oracle II") (denying motion for reconsideration of competitive bidding process used in Oracle I and approving use of bidding process to determine class counsel for claims against defendant Arthur Andersen); and

In re Oracle Securities Litig., 829 F.Supp. 1176 (N.D.Cal.1993) ("Oracle IV") (rejecting proposed class and derivative settlements).

[2] Oracle is a Delaware corporation.

[3] In McDonnell Douglas, the Delaware Supreme Court explained the so-called "mootness test":

where a stockholder's derivative suit has been rendered moot by subsequent action of the defendant, the latter has the burden of showing no causal connection between the two in order to defeat the stockholder's claim for legal fees and expenses.

310 A.2d at 637.

[4] In Allied Artists Pictures Corp., the Supreme Court of Delaware stated "[t]he main concern apparent in our cases has been that the party who takes the action that cures the alleged wrong to the corporation's benefit and thereby moots or settles the lawsuit should bear the burden of demonstrating that the lawsuit did not in any way cause their action." 413 A.2d at 880.

[5] While Sugarland Industries was a shareholder derivative suit, Six Mexican Workers was a class action. Nevertheless, the Ninth Circuit's general standards for determining attorney fees in common fund class actions articulated in Six Mexican Workers guide this case as well. The only palpable result of this derivative suit was the monetary fund of $1.75 million Arthur Andersen agreed to contribute to the class action settlement fund. See supra pp. 1446-1449. Because there were no nonmonetary benefits conferred upon Oracle by this derivative suit, common fund principles are particularly pertinent. Cf. Reiser v. Del Monte Prop., 605 F.2d 1135, 1139-40 (9th Cir.1979) ("[i]n both the common fund and common benefit [i.e., shareholder derivative] cases, the [Supreme] Court has refused to place form over substance, focusing instead on the actual effect of a plaintiff's suit"). Accordingly, the court will look to fee approaches used in common fund cases for guidance in determining derivative counsel's fees.

[6] The court is not aware of any Ninth Circuit case mandating standards for determining derivative counsel's fees different from those governing common fund cases. The Ninth Circuit cases that do mention derivative counsel fee calculations state only that the determination of the amount of the award falls within the discretion of the district court. See, e.g., Lewis v. Anderson, 692 F.2d 1267, 1272 (9th Cir.1982).

[7] As suggested by Judge Sneed in Six Mexican Workers, 904 F.2d at 1312 (concurring opinion), and for reasons extensively discussed in Oracle I and Oracle III, neither the "lodestar" nor the "benchmark" approaches are effective substitutes for determining the fair market value of counsel's services at the outset of common fund litigation. This point emerges from the discussion of class counsel's compensation below.

[8] Misconstruing this point, one of derivative counsel argued orally that because the $750,000 cap emerged from negotiations with defendants, this number best approximates the market. But a fee-ceiling is surely not the same thing as a fee bargain.

[9] The court does not understand Ninth Circuit jurisprudence to establish unalterable lower and upper bounds for fee awards at 20 and 30 percent, respectively. But these do appear to be levels at which the Ninth Circuit has required extraordinary justification for further departures and thus may better be termed "sticking points."

[10] Allowing attorneys to keep the whole recovery would reduce incentives for them to strike settlements at amounts which "sell out" the plaintiff's claims at less than full value.

[11] It is important to note that the Dague Court did not acknowledge a global trend favoring lodestar over percentage fees. An analysis of the case that the Court cites for this "trend," Venegas v. Mitchell, 495 U.S. 82, 110 S.Ct. 1679, 109 L.Ed.2d 74 (1990), shows that the preference for lodestar only applies in the statutory fee-shifting context. The Venegas Court simply recognized that "in construing [the fee-shifting provision of 42 USC] § 1988, [courts] have generally turned away from the [percentage of recovery] contingent-fee model to the lodestar model of hours reasonably expended compensated at reasonable rates." Id. at 87, 110 S.Ct. at 1682.

[12] As of May 10, 1994, these claims processing expenses apparently total $187,312.11. See Ltr of Neil L. Selinger, May 10, 1994.

[13] $4,800,000 + $825,000 = $5,625,000; $5,625,000/$25,000,000 = .225 (= 22.5%).

[14] Lowey, Dannenberg, Bemporad, Brachtl & Selinger, P.C.

[15] Abbey & Ellis.

[16] Berger & Montague, P.C.

[17] David B. Gold, PLC.

[18] In discussing the issue of derivative counsel's fees above, the court noted the problem with enhancing percentage of recovery fee awards to compensate lawyers fully for their risk of non-recovery: too high a percentage award can engender burdensome and wasteful longshot common fund litigation. See supra pp. 1452-1453. Excessive risk multipliers in the lodestar context will similarly encourage longshot litigation. Hence, there should be no one-to-one correspondence between the risk of non-recovery and the risk multiplier used under the lodestar method of awarding counsel fees (i.e., there must be a limit to the amount that lawyers can obtain as compensation for the risks of non-recovery).

[19] The relevant period is January 3, 1990, through March 27, 1990.