ORDER APPOINTING CLASS COUNSEL
WALKER, District Judge.Four of the law firms representing Oracle shareholders1 have bid to serve as class counsel. This order describes and compares those bids, and then selects class counsel on the basis of the bidders’qualifications and fee proposals. In addition, the order modifies the method contemplated in the August 3 order for reimbursing class counsel’s out-of-pocket expenses and considers certain suggestions by counsel in the two related derivative cases2 regarding the selection of class counsel. 131 F.R.D. 688.
I. DESCRIPTION OF THE BIDS.
The court’s August 3 order required each bidder to specify “the percentage of any recovery such firm will charge as fees and costs in the event that a recovery for the class is achieved.” August 3 Order at 697 (emphasis supplied). Two bids, however, specified a percentage fee for attorney compensation only although one of these bids sets a ceiling on the amount of litigation expenses to be reimbursed from the recovery. One bidder submitted two bids, one including expenses and one excluding expenses, and requested modification of the requirement that expenses be paid from the percentage fee; expenses are too difficult to estimate, argues this bidder. Berger Bid at 4-5. Although only one bidder cast its bid as directed,3 it will become apparent that these differences do not prevent comparing the bids and selecting class counsel.4
In addition to specifying percentage attorney fees, the bids describe the bidders and propose alternate contingent events as the court permitted. See August 3 Order at 697 n. 22.
A. Abbey & Ellis.
Abbey & Ellis is located in New York City. The firm has ten lawyers, six partners and four associates, and a staff of thirteen, including three paralegals; Abbey Bid at 1-2. Arthur Abbey and Ralph Ellis, the firm’s two most senior lawyers, are seasoned litigators with extensive experience in plaintiff class action work. Id. at Exh. A.
Abbey & Ellis initially bid a straight 24% contingent fee covering both attorney fees and expenses (except expert fees), later modified to 22.5% for attorney fees alone. Spanier Letter dated September 7, 1990. The Abbey bid takes least advantage of the option to specify alternate contingent events, but to prevent “any windfall to itself at the class’ expense,” the Abbey firm does propose an unspecified downward adjustment in its fees if the litigation is settled early. Abbey Bid at 12. Beyond that, the Abbey bid proposes no specific alternate contingencies, stating that it is impossible to predict the outcome of the litigation accurately enough to do so.
*540B. Berger & Montague, P.C.
Berger & Montague, P.C., is located in Philadelphia. The firm has a relationship with a San Jose, California firm sufficient to merit a notation on its letterhead, although the bid makes no effort to describe that relationship or how, if at all, it should affect selection of class counsel. The Berger firm has 42 lawyers which together with a support staff add up to total personnel of over 115 persons. Berger Bid at 9. The firm’s bid recounts extensive experience in a wide variety of litigation, with notable emphasis on securities and antitrust cases. Id. at 10-18. The Berger lawyers appear seasoned and widely experienced. Id. at 18-26.
The Berger bid5 proposes a fairly elaborate set of alternate contingencies keyed to: (1) when the litigation is resolved, and (2) amount of recovery. The bid specifies different percentages following three stages of the litigation: (1) document discovery, (2) .deposition discovery, and (3) commencement of trial. In addition, the Berger bid proposes to surcharge the class one percent of the recovery if a settlement is achieved within one year and another surcharge in the event of appellate proceedings. The Berger bid may thus be viewed as five separate fee proposals for various levels of recovery.
Because settlement within one year would most likely find the litigation still in the first of the Berger bid’s three stages, document discovery, the bid’s early settlement proposal (“Berger #1”)6 is calculated by applying the one percent surcharge to the document discovery settlement proposal (Berger #2). Next, the Berger bid contemplates settlement after commencement of depositions, but again without the early settlement bonus (“Berger #3”), then settlement achieved after commencement of trial but without an appeal (“Berger #4”), and finally applying the appeal surcharge7 to the trial contingency yields the proposal for recovery achieved after initiation of appellate proceedings (Berger #5).
Contingencies8
Early Document Deposition
Settlement Stage Stage Trial Appeal
Recovery (Berger #1) (Berger #2) (Berger #3) (Berger #4) (Berger #5)
Up to $10M 27% 26% 27% 32% 37%
$10M-$20M 25% 24% 25% 30% 35%
$20M-$50M 25% 24% 24% 30% 32%
$50M or more 25% 24% 24% 29% 30%
*541C. David B. Gold, PLC.
David B. Gold, PLC, is located in San Francisco and has 12 lawyers. The firm, which has extensive experience in complex business and class litigation, claims to have originated the fraud-on-the-market theory recognized in Basic, Inc. v. Levinson, 485 U.S. 224, 108 S.Ct. 978, 99 L.Ed.2d 194 (1988). Much of the Gold firm’s work has involved technology-related industries. Gold Bid at 4-8. Its founder, David B. Gold, is affiliated with the Center on Conflict and Negotiation at Stanford University, an interdisciplinary research organization. Id. at 9. The Gold firm has frequently opposed defendants’ lead counsel, Melvin R. Goldman of Morrison & Foerster. Id. at 8. The Gold bid proposes a basic fee schedule, excluding expenses:
Recovery Marginal
Fee Rate
Up to $30M 25%
$30-$60M 22%
$60-$100M 20%
$100-$130M 17%
$130-$160M 14%
$160-$200M 12%
$200M or more 10%
In addition, the Gold bid proposes a declining surcharge based upon the time for recovery:
Time for Resolution Surcharge
Up to 12 months 5%
Up to 18 months 4%
Up to 24 months 3%
Up to 30 months 1%
The Gold bid’s basic fee schedule, adjusted for the surcharge, may be viewed as five separate proposals:
Time for Resolution (months)
0-12 13-18 19-24 25-30 30 or more
Recovery (Gold #1) (Gold #2) (Gold #3) (Gold #4) (Gold #5)
Up to $30M 30% 29% 28% 26% 25%
$30M-$60M 27% 26% 25% 23% 22%
$60M-$100M 25% 24% 23% 21% 20%
$100M-$130M 22% 21% 20% 18% 17%
$130M-$160M 19% 18% 17% 15% 14%
$100M-$200M 17% 16% 15% 13% 12%
$200M or more ' 15% 14% 13% 11% 10%
D. Lowey, Dannenberg, Bemporad, Brachtl & Selinger, PC.
Lowey, Dannenberg, Bemporad, Brachtl & Selinger, P.C., is located in New York City and consists of eleven lawyers, five of whom are principals. Lowey Bid at Exh. 1. The firm has a support staff of approximately twenty, including five paralegals, and specializes in class and derivative litigation in securities and corporate finance matters. On the complaint it has filed in this litigation, the Lowey firm associated a San Diego law firm, but the bid specifies no role for that firm.
The Lowey bid proposes a basic fee schedule with an early settlement discount, and a proposed cap or limit of $325,000 on the amount of litigation expenses to be charged the class. The early settlement proposal (Lowey #1) applies to any recovery achieved within 12 months and reduces by 20% the Lowey bid’s basic contingent fee proposal (Lowey #2) which applies to all recoveries thereafter:
Time for Resolution (months)
0-12 13 or more
Recovery (Lowey #1) (Lowey #2)
Up to $1M 24% 30%
$1M-$5M 20% 25%
$5M-$15M 16% 20%
$15M or more 12% 15%
*542II. COMPARISON OF THE BIDS.
A. The Bidders' Qualifications.
In its August 3 order, the court invited each bidder to set forth its qualifications and urged emphasis on the “contribution such firm made to the welfare of the class plaintiffs” in prior class actions. August 3 Order at 697. Each of the bidders has, for the most part, submitted little helpful qualitative information. Most of it consists of “celebrity endorsements”—recitations of flattering remarks by judges before whom the bidders have litigated. The remarks are subjective and contextual, basically pufferies.9
These are particularly unhelpful because, different from individual clients, most class members are unlikely ever to deal with class counsel except, possibly, through the notice process. Personality and the myriad other intangible factors which typically influence lawyer selection by individual clients are immaterial in selecting class counsel.10
All of the bidders are prominent, well established firms specializing in the type of litigation at bar. All appear adequately qualified to serve as class counsel, and it is impossible objectively to distinguish among these firms in terms of their background, experience and legal abilities. One qualitative distinction does, however, emerge. The Gold and Lowey bids each establish a qualitative advantage over the other bids by an emphasis on minimizing litigation expenses.
The Gold bid points to locational factors which suggest litigation economies: (1) San Francisco location should minimize expenses of court appearances and deposition discovery as most witnesses are located nearby; and (2) extensive past litigation experience with and proximity to defendants’ lead counsel may facilitate resolution.11
The Lowey bid establishes a qualitative advantage by committing to cap litigation expenses charged to the class at $325,000. The potential value of this commitment is suggested by the other bids. The Gold bid estimates that litigation expenses could range from $560,000 to $880,000. Gold Bid at 11. The rejected Berger-Gold joint proposal estimated pretrial expenses at $460,-000 to $780,000, with an additional $400,000 to $600,000 necessary in the event of trial.12 These estimates confirm that the Lowey bid’s expense cap offers significant protection against depletion of the common fund through unchecked litigation expenses and is, therefore, a significant qualitative factor favoring that bid.
Qualitative superiority of the Gold and Lowey bids does not, however, end the inquiry. Quality must warrant its cost.
B. Recovery Contingencies.
Competitive bidding has elicited from three of the bidders a factor notably lacking in the judge-devised benchmark percentage fee suggestions of In re Activision Secur. Litigation, 723 F.Supp. 1373 *543(N.D.Cal.1989) and Paul, Johnson, Alston & Hunt v. Graulty, 886 F.2d 268 (9th Cir.1989). The Berger, Gold and Lowey bids propose variable percentage fees, all of which decline as recovery increases. Only the Abbey bid seeks a constant percentage fee. A declining percentage is readily explained: increasing amounts of recovery do not require correspondingly increased levels of attorney effort. To make their bids more competitive, the three bidders offer to share these economies of effort with the class.
The declines in the percentages of these bids, however, are not proportionate to the changing levels of recovery. This means that for higher recoveries, higher attorney fees are generated. But the recoveries and fees increase at different rates. This differential preserves the incentives for lawyers inherent in percentage fees while at the same time addressing the enduring criticism that they lead to attorney windfalls. Judges, less attuned than lawyers to the economics of law practice, are unlikely to know where to place the breakpoints or how to adjust the rates to achieve this balance of incentive and windfall protection.13 Competitive bidding has drawn the bidders here to seek that balance and, thereby, share the benefits of their economies of effort with the class. See W. Lynk, “The Courts and the Market: An Economic Analysis of Contingent Fees in Class-Action Litigation” 19 J. Legal Studies 247, 255 (1990).
Among the three bids, the percentage reductions vary considerably and are spread widely over a range of recoveries. The Gold and Lowey percentages decline fairly sharply, going from maxima which are at least twice their corresponding mini-ma. Gold #1 and Lowey #2, for example, go from 30% to 15%. Gold #4 and #5 have even sharper rates of decrease, minima in the latter being only 40% of their corresponding maxima. The percentage reductions in the Berger bid are more gradual. Berger #1 goes only from 27% to 25% while the most deeply declining Berger proposal (#5) goes only from 37% to 30%.
The ranges of recovery over which the bidders have spread their percentage reductions also differ markedly. The Lowey bid covers a short range, achieving minima at $15M, while the Gold proposals do not reach their minima until a recovery of $200M. The relatively shallow Berger reductions touch bottom at increasing recovery amounts, going from $10M for Berger #1 and #2, $20M for Berger #3, $50M for Berger #4 and #5.
These percentage reductions have a notable effect on the class’ net recovery, a point made evident by comparing the variable percentage bids to the Abbey straightline. The Gold bid includes some percentages lower than the Abbey constant, but none of these applies to any recovery under $30M. See, e.g., Gold #5. Indeed, the percentage reductions in the Gold bid cover such a large range of recoveries that even its lowest proposal (Gold #5) would not generate lower fees (and thus higher net recovery for the class) than the Abbey bid until a recovery of $84M were achieved. Lowey #1, on the other hand, achieves the same or greater net to the class as the Abbey constant beginning at a $1.6M recovery. Lowey #2 achieves parity with Abbey at $12M without regard to the effect of the $325,000 expenses cap. Significantly, no fee proposed in the Berger bid at any level of recovery affords as high a net recovery to the class as the Abbey bid.
Although generally endorsing percentage fees, derivative counsel stress the agency problem inherent in that form of lawyer compensation: percentage contingent fee lawyers have an incentive “to settle prematurely and cheaply.” Deriva*544tive Counsel Submission at 5.14 To address this problem, derivative counsel urge not a percentage fee which decreases with larger recoveries, but one which increases with larger recoveries. Such a percentage fee was adopted in the Lincoln Savings litigation:
Total Recovery Fees
$0—$150M 25%
over $150M 29%
This compensation scheme, in the undersigned’s view, is unsatisfactory.
The first difficulty with the increasing percentage approach is its obvious discontinuity. In the Lincoln Savings litigation, for example, no recovery less than $158.45M is worth as much to the class as a recovery of $150M. Second, an increasing percentage fee deprives the class of the economies of effort implicit in larger recoveries, something which competition forced the lawyers here to share. Third, and perhaps most seriously, an increasing percentage fee mixes too many variables. The “sell-out settlement” or “underinvestment” problem to be corrected is insufficient attorney effort to realize the optimal level of class recovery. But an increasing percentage rewards additional recovery, not additional effort. Amount of recovery is thus used (perhaps inadvertently) as a surrogate for attorney effort.15 But amount of recovery in litigation does not equate with the amount of effort necessary to realize it.
Recovery reflects many factors other than attorney effort. These include quality and availability of evidence and witnesses, state of the law and willingness of the defendant to fight, among other things. Increasing the percentage of class counsel’s share of the recovery could thus produce attorney windfalls unrelated to the quality or quantity of attorney services provided. One can easily imagine a scenario in which the increasing percentage fee arrangement actually discourages, not encourages, additional attorney effort and thus aggravates the agency problem. It is, perhaps, a complete answer to derivative counsel’s suggestion of an increasing percentage fee that none of the bidders—who, bafter all, are putting their own effort and money on the line—advanced such a bid.16
C. Time and Event Contingencies.
The absence of any supervision by clients has long been recognized as a source of the agency problem in class actions. See, e.g., H. Kalven and M. Rosenfield, “The Contemporary Function of the Class Suit,” 8 U.Chi.L.Rev. 684 (1941).17 The classic manifestation of the problem in a class action involves a non-pecuniary settlement (e.g., injunctive relief), “expert valued” at some fictitious figure, together with arrangements to pay plaintiffs’ lawyers their fees. The defendants thus get off cheaply, the *545plaintiffs’ (and defendants’) lawyers get the only real money that changes hands and the court, which approves the settlement, clears its docket of troublesome litigation. Little wonder that “[a]ll the dynamics conduce to judicial approval of such settlements.” Alleghany Corporation v. Kirby, 333 F.2d 327, 347 (2d Cir.1964) (Friendly, J., dissenting), affirmed en banc, 340 F.2d 311 (2d Cir.1965), cert. dismissed, 384 U.S. 28, 86 S.Ct. 1250, 16 L.Ed.2d 335 (1966).
Even settlements in which large sums are paid to the class may be collusive, a tendency possibly reinforced by lodestar attorney compensation. Coffee, “Understanding the Plaintiff’s Attorney: The Implications of Economic Theory for Private Enforcement of Law Through Class and Derivative Actions,” 86 Col.L.Rev. 669, 717-718 (1986). Eliminating lodestar compensation in common fund cases may ameliorate the problem, Id. at 716 n. 129, but affords no panacea because, as previously noted, percentage fee arrangements contain an agency problem wholly apart from the class feature. See n. 14, supra. In simplest terms, a lawyer who shares only a part of the victim’s recovery has less incentive than the victim to devote the effort necessary to achieve the level of recovery which is optimal for the victim. Consequently, the lawyer may be all too willing to “sell-out” (i.e., settle the client’s claim for less than it is worth to the client).
While it is by no means clear that the bidders here consciously attempted to deal with this agency problem, certain features of their bids bear directly on that subject. All the bids, for example, make some provision for an early settlement. The Abbey bid does not spell out any details, offering only some unspecified reduction in fees in the event of an early (again unspecified) settlement. Abbey Bid at 12. The Lowey bid specifically provides that if settlement were achieved within 12 months, that firm’s proposed contingent fee would be reduced by 20 percent. Lowey Bid at 4.
The Berger and Gold bids spell out a different response to early settlement. If resolution is achieved within 12 months, the Berger bid would surcharge the class an additional one percent while the Gold firm would surcharge five percent of the recovery. The Berger and Gold early settlement proposals thus suffer from the same discontinuity and covariance problems noted in connection with the Lincoln Savings fee award. Supra at 544. More seriously, the Berger and Gold bids actually reward the lesser attorney effort associated with an early settlement. Although a “reward” for early resolution is confined to the early settlement feature of the Berger bid, this concept underlies the whole structure of the Gold bid. As the time needed to achieve a resolution of the litigation mounts, the Gold firm’s compensation diminishes.
While not articulated in either the Berger or Gold bids, the early settlement surcharges and the declining percentage approach of the Gold bid can be explained only by an incentive rationale: better results should beget a bigger fee. The underlying premise must be that higher percentage fees to the lawyers will call forth the effort needed to achieve a recovery sooner rather than later. Derivative counsel offer this explanation, Derivative Counsel Submission at 5-6, but it cannot withstand scrutiny.
After the Sturm und Drang of the August 3 order, it may seem anomalous to reprove a performance linked incentive approach to attorney compensation offered by the lawyers themselves. To accept derivative counsel’s rationale and adopt the early settlement features of the Berger and Gold bids, however, would depart from an important principle. Attorney fees in this type of case are paid as “ ‘reasonable compensation for creating a common fund.’ ” State of Florida v. Dunne, 915 F.2d 542, 545 (9th Cir.1990). The lodestar approach is grounded on this principle and this court, while seeking to avoid the practical difficulties of lodestar computations, must build its alternative method from the same rock. Ibid. Percentage fees which increase due to early settlements (Berger and Gold early settlement surcharges) or greater recoveries (Lincoln Savings) are incompatible with a reasonable compensation theory. At its limit, the rationale of *546these “incentives” would justify awarding the entire recovery to the lawyers and none to the class for only when the lawyers own the whole claim will their incentives be identical to those of the class. But this is a formula for converting class action lawyers into private parens patriae.18 See generally, Hawaii v. Standard Oil Co., 405 U.S. 251, 257, 260, 92 S.Ct. 885, 890, 31 L.Ed.2d 184 (1972). The court finds no legal policy to elevate lawyers to quasi-sovereigns.19
Conversely, the presumed disincentives of percentage fees which decline with time, as in the Gold bid, also are inconsistent with redress of the class’ injury. Because the passage of time diminishes class counsel’s compensation under the Gold bid, the temptation of counsel to succumb to a “sell-out settlement” would be aggravated, not diminished. Rather than put in the additional effort to increase the class’ recovery, counsel under the Gold bid could improve its take by the earliest possible settlement even though spending additional time and effort could improve the class’ recovery.
Apart from its early settlement surcharges, the Berger bid’s other time related contingencies afford incentives consistent with the concept that class counsel are entitled to “reasonable compensation.” The Lowey bid does as well. The Berger and Lowey proposals do so by providing increased compensation for increased attorney effort (in the Berger bid, of course, only after one year). The Lowey bid provides for 25 percent higher compensation levels if the resolution takes longer than 12 months.20 For recoveries under $10M, the Berger bid steps up basic compensation levels five percent after commencement of depositions, four percent after start of trial and five percent upon appeal. The greater compensation called for by these specified events is warranted by greater attorney effort21.
For reasons well articulated in the lodestar context, Blum v. Stenson, 465 U.S. 886, 104 S.Ct. 1541, 79 L.Ed.2d 891 (1984), a reasonable fee is one “calculated according to the prevailing market rates in the relevant community.” Id. at 895, 104 S.Ct. at 1547. A reasonable percentage fee is also one which reflects market forces. Such a percentage fee is likely to be one in which “class-action plaintiffs' lawyers obtain a smaller fraction of the total recovery the larger the recovery is, and a greater fraction of the total recovery the longer they must wait to be paid.” W. Lynk, “The Courts and the Market: An Economic Analysis of Contingent Fees in Class-Action Litigation,” 19 J.Legal Stud. 247, 258 (1990).22
*547Of the four bids, only the Berger and Lowey bids conform to the fee structure associated with a competitive market, and the former only after the first year of litigation. But selecting the “fee structure an informed, sophisticated client would use to compensate his attorney when close monitoring is not feasible,” Coffee, “Understanding the Plaintiffs Attorney: The Implications of Economic Theory for Private Enforcement of Law Through Class and Derivative Actions,” 86 Col.L. Rev. 669, 697 (1986), is not all there is to selecting the right class action lawyer. The level of fees is also important and, in this regard, Lowey trumps Berger. Only the Abbey bid is competitive in rates to the Lowey bid, but the Abbey bid provides no protection against runaway litigation expenses, its early settlement discount is left wholly to speculation and its constant percentage exposes the class to the danger of attorney windfalls. The Gold bid, while offering a limited qualitative superiority to all but the Lowey bid, is misstructured and the levels of its fees uncompetitive except at truly extraordinary levels of recovery.
The practical question is whether the Lowey bid is “too good.” Does it provide an insufficient level of attorney compensation, so that the Lowey firm will be irresistibly tempted to do slipshod work or “sellout” the class? Several factors suggest a negative answer. First, although the court approval requirement of Fed.R.Civ.P. 23 affords little resistance to the “conducing dynamics” toward settlement approval, the potential adverse reputational consequences to the Lowey firm of proposing a “sell-out settlement” are not negligible. See P. Danzón, “Contingent Fees for Personal Injury Litigation,” 14 Bell J. Econ, 213 (1983). Because the Lowey firm and other bidders are “repeat” players in securities class actions, announcement of the court’s concern for “sell out settlement” potential at the outset of litigation may help to guard against it. Id. at 217. Second, the main claims at bar being Federal, a presumption favoring the award of prejudgment interest at defendants’ cost of funds applies. Gorenstein Enterprises, Inc. v. Quality Care-USA, Inc., 874 F.2d 431, 436 (7th Cir.1989), disagreed with on other grounds, Business Guides, Inc. v. Chromatic Communications Enterprises, Inc., 892 F.2d 802 (9th Cir.1989). Such an award eliminates the penalty to the class and class counsel implicit in delay, whether induced by defendants, the court or otherwise. This mitigates class counsel’s temptation to accept a “sell-out settlement.” Coffee, supra, 86 Col.L.Rev. at 725.
Finally, the Lowey bid is not out of line with that produced by other fee regimes. In fact, if after more than one year (and thus subject to Lowey #2) this litigation were to settle for the same amount as did Activision, the net attorney fees to the Lowey firm would be $4,000 more than Judge Patel awarded the attorneys in that case under the lodestar. In re Activision Secur. Litigation, 723 F.Supp. 1373, 1379 (N.D.Cal.1989).23 If such a settlement were early enough to be subject to Lowey #1, then of course the attorney fees here would be much lower than in Activision. If the settlement here were significantly larger than in Activision, then too the proportion of attorney fees and litigation expenses here would be lower, and net class recovery higher, than under the 30% benchmark announced in Activision, which was the basis of the rejected Berger-Gold joint proposal. Unlike constant percentage fees, the Lowey bid is geared to a variety of outcomes and will generally afford greater net recovery for the class (especially at larger recoveries), but appears to provide compensation consistent with amounts courts have found appropriate.24
The court is satisfied that competitive bidding here has produced serviceable class counsel arrangements fully consistent with the Ninth Circuit’s recently re-affirmed *548standard of reasonable compensation under the circumstances. State of Florida v. Dunne, 915 F.2d 542, 546 (9th Cir.1990). Moreover, competitive bidding has at the outset of the litigation, rather than at its end, fulfilled the requirement to assess the reasonableness of the fee application by reference to other applications. Ibid. The bid selected offers the class lawyers performance-related incentives while also protecting against lawyer windfalls. These have been accomplished without the “protracted, complicated, and exhausting” fee litigation that typically accompanies lodestar determinations. Pennsylvania v. Delaware Valley Citizens’ Council for Clean Air, 483 U.S. 711, 722, 107 S.Ct. 3078, 3085, 97 L.Ed.2d 585 (1987).
III. LEAD COUNSEL SELECTION.
For the foregoing reasons, the court designates as lead class counsel herein, the law firm of Lowey, Dannenberg, Bemporad, Brachtl & Selsinger, P.C., in accordance with the fees and costs proposal submitted to the court under letter dated August 20, 1990 which is ordered filed in the records herein together with the proposals of the competing firms. The twelve month period specified in the Lowey firm’s August 20 letter bid shall commence with the filing of this order.
Lead class counsel shall contact all other counsel regarding the earliest convenient date for further status conference.
. Abbey & Ellis; Berger & Montague, P.C.; David B. Gold, PLC; Lowey, Danneberg, Bemporad, BrachtI & Selinger, P.C.
. Leiff, Cabraser & Heiman and Kohn, Savett, Klein & Graf, P.C.
. The court requested and received a modified attorney fees-only bid by that firm, Abbey & Ellis, on September 7, 1990.
. The litigation expenses referred to here are those other than attorney fees and costs recoverable under Rule 54(d) in the event the class prevails at trial. See C. Wright, et al., Fed.Prac. and Proc., § 2666, at 173-174 (1983 ed.). Litigation expenses can materially affect the “payoffs” or incentives for class litigation. See, generally, D. Dewees, et al., "An Economic Analysis of Cost and Fee Rules for Class Actions,” 10 J. Legal Studies 155 (1981). The court believes that the risk of misestimating such expenses should rest with class counsel who, after all, incur them. For reasons soon to be explained, however, the reimbursement arrangements here can be modified with little violence to this belief.
. The Berger firm asks the court to delete from its August 3 order attribution to Ms. Savett of the description of Mr. Gold as "obnoxiously old-fashioned.” Berger Bid at 1 n. 1. While this characterization of Mr. Gold was not, in fact, uttered at the May 4 hearing, the Berger firm’s memorandum for that hearing, a document signed by Ms. Savett, so describes Mr. Gold. Milberg and Berger Reply Memorandum filed April 27, 1990, at 17. The moral of recounting counsels’ May 4 "shouting match” is made more pointed by the fact that their competitive bids lack rhetorical attacks on fellow counsel.
. Conceivably, the litigation could proceed to deposition discovery but still settle within one year, in which case the early settlement bid would be: (M = 000,000) up to f 10M—28%; $10M-$20M—26%; and $20M or more—25%.
. The Berger bid includes a typographical error that the one percent surcharge applies to any recovery exceeding $20M, but the court made the necessary correction. See Berger Bid at 8.
. The Berger bid including litigation expenses within the percentage fee contains the same contingencies, but adjusts upward the share of recovery demanded:
Recovery Berger Berger Berger Berger Berger
#1A #2A #3A #4A #5A
Up to $10M 28% 27% 32% 36% 41%
S10M-S20M 26% 25% 26% 33% 38%
$20M-$50M 25% 24% 24% 33% 35%
$50M or more 25% 24% 24% 30% 31%
. Such judicial bouquets are also problematic credentials. They suggest high ability, but this comes at a price. The bidders’ reputations also suggest that there is a high opportunity cost to their time, which may correlate inversely with their willingness to put effort into the litigation and hence with the probability of winning. W. Lynk, "The Courts and the Market: An Economic Analysis of Contingent Fees in Class-Action Litigation,” 19 J. Legal Stud. 247, 255 (1990). This effect is not a factor in the court's selection, however, since all bidders were highly praised.
. The Gold firm urges that the bids be evaluated solely on the bidders’ reputations and professional standing, Gold Bid at 2, because "you get what you pay for.” Id. at 4. Unquestionably, for example, the class would be better served by an $11M recovery in 20 months by the Gold firm than a similarly timed $10M recovery by the Abbey firm, assuming expenses are the same. Compare Gold #3 to Abbey. But reliance solely on qualitative factors is unwarranted here because none of the bidders has demonstrated qualitative distinctions sufficient to outweigh price considerations.
. Mr. Gold’s association with the Stanford Center on Conflict and Negotiation is an intriguing but incalculable credential. See Gold Bid at 8-9.
. Comparing the expenses-included and -excluded Berger bids at their projected recovery levels suggests that firm believes expenses could exceed $500,000. Berger Bid at 6-9.
. The undersigned is aware of two recent instances in which courts have felt compelled, on their own initiative, to cut the percentage contingent fees of certain lawyers to avoid unfair results. In Re Johns-Manville Corporation, et al. Reorganization Proceedings, Nos. 82 B 11656— 82 B 11676, 1990 WL 123830 (E.D. and S.D.N.Y. Aug. 21, 1990) and Gagnon v. Shoblom, No. 88-2105 (Mass.Super.Ct., undated) appeal pending. See Wall Street Journal, Sept. 19, 1990 at Bl. If the lawyer markets had been competitive, such drastic measures might have been avoided. See August 3 Order at 696 n. 20.
. This is the problem described by Judge Posner in Chesny v. Marek, 720 F.2d 474 (7th Cir.1983), reversed on other grounds, 473 U.S. 1, 105 S.Ct. 3012, 87 L.Ed.2d 1 (1985), and pointed out by this court. August 3 Order at 694 n. 16. The problem commonly goes by the name "sellout” or, in the class context, "sweetheart” settlement, J. Coffee, "The Regulation of Entrepreneurial Litigation: Balancing Fairness and Efficiency in the Large Class Action," 54 U.Chi.L. Rev. 877, 883 (1987). The bar probably would prefer a more euphemistic term, such as “underinvestment in litigation." See, e.g., Derivative Counsel Submission at 7. But contingent fee litigation and class actions are not distinct as all types of attorney fee arrangements pose an agency problem of some type. August 3 Order at 694.
. No where is this confusion more apparent than in derivative counsel’s own submission. Derivative Counsel Submission at 8-10. Counsel attempt to depict the Lincoln Savings fee arrangement along the lines of a graph in J. Coffee, "Understanding the Plaintiff's Attorney: The Implications of Economic Theory for Private Enforcement of Law Through Class and Derivative Actions,” 86 Col.L.Rev. 669 (1986). But in doing so, derivative counsel have arrayed dollars on the ordinate and months of litigation on the abscissa, despite the fact that the latter variable is nowhere to be found in the Lincoln Savings formula.
. An economic model of contingent fees confirms that recovery and contingent fee percentage should be inversely related under competitive conditions. W. Lynk, "The Courts and the Market: An Economic Analysis of Contingent Fees in Class Action Litigation” 19 J.Legal Stud. 247, 255, 258 (1990).
. An excellent description of the agency problem in the class action context appears in A. Rosenfield, "An Empirical Test of Class Action Settlement," 5 J.Legal Stud. 113 (1976).
. The notion has been mentioned in passing. H. Kalven and M. Rosenfield, "The Contemporary Function of the Class Suit,” 8 U.Chi.L.Rev. 684, 718, n. 98 (1941); A. Rosenfield, "An Empirical Test of Class Action Settlement,” 5 J.Legal Stud. 113, 117 (1976). At least one variant calling for recovery by a non-profit self-regulatory body has respectable support, see, e.g., Am.Law Inst., Fed.Securities Code, Pt. XVII § 1711 at 744 (1980), which embodies a "theory of compensation if practicable but in any event deterrence and avoidance of unjust enrichment.” The idea may also arise if "the costs of actually effecting compensation to the members of a numerous class may be extremely high, and in some cases may exceed the benefits in deterrence yielded by the action," so that requiring compensation may diminish the allocative efficiency of a class action. Posner, Economic Analysis of Law at 537 (3d ed.1986). These speculations aside, victim compensation is the touchstone of our law.
. Even when, in the 1976 amendments to the Clayton Act, Congress provided for parens patriae recovery by state attorneys general, it made specific provision "to afford each [injured] person a reasonable opportunity to secure his appropriate portion of the net monetary relief." 15 U.S.C. § 15e.
. The Abbey bid may also be said to do so by its anti-windfall provision, but because its terms are not specified, the curative effect is essentially speculative.
. This aspect of the Berger and Lowey proposals displays a discontinuity similar to that noted in the Lincoln Savings plan. Under the Lowey bid, for example, a $4M settlement within 12 months is worth as much to the class as a $4.28M settlement which takes more than one year. Because the Berger bid contains more contingencies, it would exhibit more such discontinuities.
. Contrary to Activision, supra, at 542-43, empirical evidence negates the assumption of a "natural” or generally accepted percentage fee, whether judicially or market determined. W. Lynk, supra, at 259 n. 23.
. The attorneys in Activision were paid $1,161,-413.81 and expenses of $398,705.55 were deducted from the class recovery; these totalled 32.8% of the recovery. Given a similar recovery, Lowey #2 would produce fees of $1,165,409.81 (net of expenses deducted from the fee award) and expenses of $325,000.
. Because of its expenses cap, the Lowey bid gives class counsel real incentive to avoid unnecessary litigation expenses, consistent with the court’s initial intent in the August 3 order. See n. 4, supra.