In Re HPL TECHNOLOGIES, INC SECURITIES LITIGATION

366 F. Supp. 2d 912 (2005)

In Re HPL TECHNOLOGIES, INC. SECURITIES LITIGATION

No. C-02-3510 VRW.

United States District Court, N.D. California.

April 22, 2005.

Ilona Kohn, New York City, Robert A. Jigarjian, San Francisco, CA, Robert S. Green, San Francisco, CA, Darren J. Robbins, Lerach Coughlin Stoia & Robbins LLP, San Diego, CA, for Plaintiff.

Christopher A. Patz, Laurence Andrew Weiss, Louis Pugh, Margaret L. Wu, Raymond H. Sheen, Michael F. Tubach, Erin Elizabeth Schneider, Rohit K. Singla, San Francisco, CA, Jacqueline D. Molnar, Menlo Park, CA, Jeffrey R. Krinsk, San Diego, CA, Andrew L. Barroway, Evan Jason Smith, Stuart L. Berman, Bala Cynwyd, PA, for Defendant.

Patrick J. Coughlin, Willow E. Radcliffe, Joseph M. Barton, Steven O. Sidener, San Francisco, CA, for Movant.

ORDER

WALKER, Chief Judge.

By an order and judgments filed March 11, 2005 (Doc. 164-166), the court granted final approval to two settlements and a plan of allocation in this securities fraud class action. In the same order, the court reserved decision on lead plaintiff's motion for an award of attorneys' fees and expenses to lead counsel. Doc. # 164 at 9-12. The award of fees and expenses is to be paid out of a common fund of $17 million and 7 million shares of HPL Technologies common stock ("HPL stock") — the combined settlement consideration from all defendants. Under the proposed fee award, lead counsel would receive an award of 15% of the common fund (taken in both cash and stock), plus lead counsel's reasonable expenses. The requested 15% award translates to $2.55 million and 1.05 million shares of HPL stock.

At first glance, such a percentage seems reasonable because fee awards in other cases have captured a larger portion of common fund recoveries (although in still other cases, lower percentage fee awards have been made). Furthermore, the fee here is below what has, in some decisions, been characterized as the 25% "benchmark" for common fund fee awards. See Paul, Johnson, Alston & Hunt v. Graulty, 886 F.2d 268, 273 (9th Cir.1989).

For the reasons set forth here, the court rejects reliance solely on a comparison of the percentage fee requested here with other percentage awards or with a so-called benchmark percentage. Merely comparing percentages overlooks the factors that may make a certain percentage fee reasonable in one case and unreasonable in another. And "a theoretical construct as flexible as a ``benchmark' seems to offer an all too tempting substitute for the searching assessment that should properly be performed in each case." Goldberger v. Integrated Resources, Inc., 209 F.3d 43, 52 (2d Cir.2000). Moreover, the court has been admonished that "the benchmark percentage should be adjusted, or replaced by a lodestar calculation, when special circumstances indicate that the percentage recovery would be either too small or too large in light of the hours devoted to the case or other relevant factors." Six (6) Mexican Workers v. Arizona Citrus Growers, 904 F.2d 1301, 1311 (9th Cir.1990). So the circumstances that require adjusting the percentage need to be considered.

Percentage awards are not meant to be a windfall for class counsel at the expense of the class. Nor should "[a] lawyer's fee * * * be likened to a case of salvage, where reward is given to the successful finder, * * * with little regard to how much or how little effort the finder expended." Klein ex rel SICOR, Inc. v. Salvi, 2004 WL 596109 at *11 (S.D.N.Y.). An inspection of the circumstances in this case indicates that the requested fee is "too large in light of the hours devoted" and "other relevant factors." This conclusion — based on the circumstances and facts of this case — is bolstered by the court's review of the literature which suggests that percentage-based fees in common fund class actions systematically exceed lodestar-based fees. See, e.g., Theodore Eisenberg & Geoffrey P Miller, Attorney Fees in Class Action Settlements: An Empirical Study, 1 J Empirical Legal Stud 27 (2004) (concluding that, controlling for other variables, lodestar-based fees in common fund class actions are about 89% as much as percentage-based fees).

The court can envision no defensible normative reason in this case — or indeed in common fund cases generally — that the amount of the fee ought to depend on the method used to compute it. Both methods should result in a "reasonable" fee, and reasonableness cannot logically depend on whether the fee is expressed as a percentage of the recovery or the product of hours and rates. Hence, when counsel apply for a fee award on a percentage basis, the requested award should approximate the fee counsel would have claimed on a lodestar basis. The two fee computations can be compared by a multiplier that, in the context of lodestar fee awards, is thought to represent a premium on counsel's services reflecting the ex ante risk of taking the case and the superior results achieved in the face of that risk.

Stated another way, a proposed percentage fee can (and, for the reasons hereafter explained, should) be compared against the fee that lead counsel would have been awarded on a lodestar basis. If the multiplier implied by that comparison is reasonable, then the percentage-based fee request is reasonable as well; if the implied multiplier is unreasonably high, then so is the proposed percentage fee award. See, e.g., In re General Motors Corp. Pick-Up Truck Fuel Tank Prods. Liability Litig., 55 F.3d 768, 820-21 n. 40 (3d Cir.1995) (Becker, J) (describing the lodestar cross-check); Vizcaino v. Microsoft Corp., 290 F.3d 1043, 1050-51 (9th Cir.2002) (approving a district court's use of a lodestar cross-check).

A lodestar cross-check of this type is now in common use in district courts elsewhere in the country. See, e.g., In re Cendant Corp. Securities Litigation, 109 F Supp 2d 285, 302 (D.N.J.2000) ("Traditionally, the ``appropriate' percentage [fee award] is * * * subjected to a cross-check."), vacated and remanded by In re Cendant Corp. Litigation, 264 F.3d 201 (3d Cir.2001); In re Bristol-Myers Squibb Securities Litigation, 2005 WL 447189 at *3 (S.D.N.Y.) (citing Goldberger, 209 F.3d at 50) ("Typically, courts utilize the percentage method and then ``cross-check' the adequacy of the resulting fee by applying the lodestar method."). The Manual for Complex Litigation (Fourth) endorses the lodestar cross-check. Manual for Complex Litigation (Fourth) § 14.122 (FJC, 2004) ("The lodestar is at least useful as a cross-check on the percentage method by estimating the number of hours spent on the litigation and the hourly rate * * *."); id § 21.724. The court heeds these authorities.

To that end, the court requested at the final approval hearing on February 24, 2005, a lodestar computation from lead counsel. Lead counsel complied the very next day. Barton Decl (Doc. # 162). Upon review of the Barton declaration, which establishes a base lodestar fee of approximately $900,000, the court requested more detailed information about the breakdown of hours and billing rates of the attorneys at lead counsel's firm. Because it was deferring decision on the attorney fee issue, the court also requested further detail about the proposed award of expenses. Lead counsel has satisfied both of the court's requests, Sidener Decl (Doc. # 167), and lead plaintiff's motion for an award of attorneys' fees and expenses (Doc. # 142) is now ripe for decision. These inquiries, and class counsel's quick responses, greatly aided the court's consideration of what information is necessary to make a lodestar cross-check work. It is apparent that a fairly systematic, although not terribly difficult assessment of the time devoted to a case and the value to be attached to that time is required to enable a lodestar cross-check of a percentage fee to comply with the Federal Rules' mandate of reasonableness. See FRCP 23(h).

I

This case is governed by the Private Securities Litigation Reform Act of 1995 (PSLRA). Accordingly, the court must first determine what discretion (if any) it has under the PSLRA to depart from an award of fees and expenses that is proposed by a lead plaintiff. There is apparently no Ninth Circuit authority on whether the court retains its usual authority under FRCP 23(h) to fix an award of attorneys fees and expenses in a common fund class action. Lead counsel have not contended that the court lacks its usual authority, apparently assuming that the court possesses such authority. The court has nevertheless considered the matter.

Here, lead plaintiff did not negotiate a fee arrangement with lead counsel immediately after lead plaintiff was selected. (By "fee arrangement," the court refers to an agreement, such as an individual plaintiff would make with an attorney, that provides for compensation for the attorney's work and an allocation of the out-of-pocket expenses of litigation.) Rather, Frederick Stanske (a Senior Vice President, Portfolio Manager and board member of lead plaintiff) "personally approved of [l]ead [c]ounsel's fee request of fifteen percent (15%) of the settlement recovery." Stanske Decl (Doc # 128) ¶ 8. Lead plaintiff has not weighed in on the proposed award of expenses, and no further evidence on lead plaintiff's negotiation for or approval of the requested fee is before the court. Such a superficial fee arrangement is apparently permitted under the PSLRA, which with respect to fee arrangements requires only that the lead plaintiff "shall, subject to the approval of the court, select and retain counsel to represent the class." 15 USC § 78u-4(a)(3)(B)(v) (emphasis added). The PSLRA says nothing about when retention must occur; nor, for that matter, does "retain[ing] counsel" necessarily involve concluding a fee arrangement.

As a matter of first principles, the earlier a fee arrangement is concluded between lead plaintiff and lead counsel, the more deference the court should pay to that fee agreement. This is because fee agreements set early in the litigation — ex ante, so to speak — are more likely than ex post fee agreements to be the product of market forces (i.e., competition among counsel proposing to represent the class). See, e.g., In re Synthroid Marketing Litigation, 264 F.3d 712, 718-19 (7th Cir.2001) (Easterbrook, J). Those same market forces are thought to result in reasonable fee agreements between attorneys and clients in individual (i.e., non-class) cases. The PSLRA's lead plaintiff provisions are largely built around an ideal of private ordering and client-driven class-action litigation; it is therefore plausible that the PSLRA implicitly counsels deference to fee arrangements concluded early in the litigation. There is also Judge Shadur's wise observation that it is inappropriate — indeed, downright unfair to class counsel — to take a fee arrived at by a market-based mechanism (such as by a court-administered auction or by arms' length negotiation between lead plaintiff and lead counsel) and subject it to further review by the court; such a process will often leave class counsel with the worst of both worlds. See In re Comdisco Securities Litigation, 150 F Supp 2d 943, 947-49 (N.D.Ill.2001).

But deference to lead plaintiff is unwarranted here because of the lack of evidence that lead plaintiff and lead counsel negotiated a fee agreement early in this litigation in a way that reflects the market value of lawyer services. Rather, lead plaintiff's involvement seems to have been confined to an endorsement of lead counsel's proposed fee. (Of course, lead counsel's desire to secure that endorsement may have constrained the fee requested here, but the court has no evidence of that before it.)

In any event, even if lead plaintiff had shopped around for lawyer services and concluded a market-based fee arrangement, there is no textual indication in the PSLRA that it supplants FRCP 23(h), which obligates the court to evaluate and approve all fee awards in class actions. If anything, certain provisions of the PSLRA affirmatively command the court to remain involved in approving lead counsel and lead counsel's fees. See 15 USC § 78u-4(a)(3)(B)(v) (requiring that lead plaintiff "shall, subject to the approval of the court, select and retain counsel"); 15 USC § 78u-4(a)(6) ("[A]ttorneys' fees * * * awarded by the court * * * shall not exceed a reasonable percentage [of recovery]."). The Ninth Circuit has recognized this in dictum:

[A] lead plaintiff's retainer agreement is far from the final word on what counsel will actually get paid, because class counsel must first be appointed, "subject to the approval of the court," 15 USC § 78u-4-(a)(3)(B)(v), and in the normal case (virtually the universal case) where there is a settlement, the court must approve the actual fees paid, subject to the [PSLRA's] limitations. See 15 USC § 78u-4(a)(6).

In re Cavanaugh, 306 F.3d 726, 733 (9th Cir.2002).

Likewise, the Third Circuit's attorney fee opinions arising out of the Cendant Corp securities litigation recognize that even under the PSLRA courts must still play the central role in making fee awards. In In re Cendant Corp. Litigation, 264 F.3d 201, 285 (3d Cir.2001) (Becker, CJ), the court rejected the district court's use of an auction to select class counsel and remanded for consideration of the presumptively reasonable fee arrangement negotiated by lead plaintiff with lead counsel. Yet the court noted that the fee under the retainer agreement was "staggering in [its] size," and mandated that "the possibility of rebuttal of the presumption of reasonableness" attaching to the negotiated fee arrangement "must be seriously considered * * * on remand." Id. Such an evaluation would be made "according to the standards * * * previous cases have set down for class actions not governed by the PSLRA." Id. And the most recent Cendant opinion explained:

[W]hile the PSLRA certainly represents a shift toward the traditional attorney-client relationship, it has not wholly adopted that paradigm. Securities class actions are still class actions, and the court retains the power to award fees. See Fed R Civ P 23(h) ("In an action certified as a class action, the court may award reasonable attorney fees * * *."). And courts would be remiss if they abdicated all responsibility to the lead plaintiffs. The lead plaintiff is not the sole client in a PSLRA class action; instead, the lead plaintiff serves as a fiduciary for the entire class. A court must therefore retain oversight over lead plaintiff's compensation decisions in order to ensure that the lead plaintiff has fulfilled its fiduciary duties.

In re Cendant Corp. Securities Litigation, 404 F.3d 173, 198-99 (3d Cir.2005) (Becker, J). The court will attempt to follow this wise guidance.

The court concludes that even if the PSLRA requires deference to lead plaintiff's negotiated fee arrangement with lead counsel, no deference is owed here. Furthermore, even when deference is owed, lead plaintiff's deal with counsel must still square with objectively reasonable fees and expenses.

II

Deference or no deference, the award of expenses in this case poses no problem. Lead counsel's detailed breakdown of the expenses incurred in this case, Sidener Decl (Doc. # 167) Ex B, and their quite modest amount make it easy to conclude that counsel's expenses are reasonable. Accordingly, the court GRANTS lead counsel an award of $59,434.57 in expenses, to be paid from the cash portion of the common fund.

III

Turning to the question of a fee award, the court previously indicated that a fee of 15% of the common fund appears at first impression reasonable. Doc. # 164 at 9; Doc. # 131 at 13. In its order preliminarily approving an award of fees, the court reflected this impression:

This is a common fund class action, and counsel seeks * * * 15% of the class's gross recovery, which counsel will take in cash and HPL common stock in proportion to the settlement fund. * * * The court has been impressed with the quality of plaintiff counsel's representation in this matter, and preliminarily finds the requested fees and costs to be reasonable in light of the nature of the case, the risks involved in the litigation, the size of the recovery and the negotiation process that led to the fee agreement.

Moreover, the fee request is on the low end of the fees recovered in comparable cases. The court undertook to analyze the data compiled in Stuart J Logan, Jack Moshman & Beverly C Moore, Attorney Fee Awards in Common Fund Class Actions, 24 Class Action Rep 167 (2003), which provides data covering cases from the mid-1970s to mid-2002. Specifically, the court looked at securities class action cases in the $10 to $20 million gross recovery range, the $20 to $30 million gross recovery range and the combined $10 to $30 million gross recovery range. For those sets of cases, the range, mean and standard deviation for the percentage of the common fund devoted to attorneys' fees and costs were:

$20 to $30mm    $10 to $20mm    $10 to $30mm
Range     7.6% to 70.8%   7.6% to 55.6%   7.6% to 70.8%
Mean      27.7%           29.5%           29.0%
Std Dev   10.9%           9.0%            9.5%

The attorneys' fees and costs sought here represent, depending on the exact amount of costs and the actual value of the HPL stock, 15.3% to 15.4% of the common fund. This places the request in this case at the low end of the ranges above. In fact, assuming normally distributed fee awards, the attorneys' fees and costs sought here fall no higher than the 13th percentile of fee awards.

Doc. # 131 at 13:15-14:20.

Further consideration suggests that the court's initial impression was wrong. Application of the lodestar cross-check makes plain that this first analysis — which considered only percentage-based fees and failed to take account of the nature and amount of work put into the case — is insufficient to ensure a reasonable fee. Relying on percentages without reference to these other factors can be, like blind reliance on benchmarks, an "all too tempting substitute for the searching assessment that should properly be performed." Goldberger, 209 F.3d at 52. Here, a lodestar cross-check warrants a fee lower than that sought by lead counsel.

In applying a lodestar cross-check, the court must first determine the dollar value of the proposed percentage-based fee award. In the typical case, this is straightforward, because settlement consideration is delivered entirely in the form of cash. Here, however, a component of the settlement consideration is in the form of HPL stock. Accordingly, the court must first value the HPL stock in dollar terms.

A

While the HPL stock has a market price (since the settlement, as high as $0.93/share), that price appears something of an illusion for valuation purposes, because the market for HPL stock is highly illiquid and quite volatile. Although the court has been unable to locate volume data, the stock is traded only on the "pink sheets"; and in just the week following the issuance of the court's final approval order, the stock traded as low as $0.72/share and as high as $0.93/share. Thus, there is not only considerable uncertainty about the market price of the stock, but also the more fundamental uncertainty that lead counsel could liquidate the million-plus shares they would receive under the proposed 15% award.

Moreover, a common fund fee award serves to reward counsel for creativity and skill in enlarging a settlement fund beyond what was thought possible or likely at the inception of the case. See, e.g., Paul, Johnson, Alston & Hunt, 886 F.2d at 271 ("Since the Supreme Court's 1885 decision in Central Railroad & Banking Co of Ga. v. Pettus, 113 U.S. 116, 5 S. Ct. 387, 28 L. Ed. 915 (1885), it is well settled that the lawyer who creates a common fund is allowed an extra reward, beyond that which he has arranged with his client, so that he might share the wealth of those upon whom he has conferred a benefit."). As such, it is worth recognizing the value of class counsel's efforts at extracting a considerable amount of unconventional settlement consideration from defendants whose cash resources were limited. Treating non-cash settlement consideration as if it were cash can overstate the dollar value of counsel's proposed percentage fee and thus penalize class counsel. When non-cash consideration is valued too dearly, the percentage fee when cross-checked against a lodestar fee can make counsel a victim of their own ingenuity. Thus, for reasons of proper deterrence, effective incentives and fairness to counsel, it is appropriate to place a very modest value on the HPL stock.

Accordingly, the court concludes that it is inappropriate to value the HPL stock component of lead counsel's fee award at its market price. Instead, the court will treat the HPL stock as worth $0.20/share — a third to a fifth of its recent market price. By this valuation, the court certainly expresses no view of the stock's actual value; the court simply must place some dollar value on a large block of an illiquid security, and doing so at this very substantial discount appears fair to counsel.

B

Valuing the HPL stock at $0.20/share, the settlement fund is worth $18.4 million ($17 million in cash and $1.4 million in stock) and the court is confronted with a straightforward lodestar cross-check exercise. The first step is to establish a lodestar figure; the second step is to cross-check the proposed percentage fee against this lodestar figure.

1

Three figures are salient in a lodestar calculation: (1) counsel's reasonable hours, (2) counsel's reasonable hourly rate and (3) a multiplier thought to compensate for various factors (including unusual skill or experience of counsel, or the ex ante risk of nonrecovery in the litigation). In performing a lodestar cross-check, however, the multiplier is implied by the ratio of the proposed percentage fee to the computed lodestar fee. For example, for a proposed percentage fee of $250,000 and a corresponding lodestar fee of $100,000, the implied multiplier is 2.5. This implied multiplier may be assessed for reasonableness. Accordingly, the court need only consider counsel's reasonable hours and counsel's reasonable hourly rate in computing the lodestar.

For hourly rates, the court will simply use current (i.e., 2005) hourly rates; doing so simplifies the calculation and accounts for the time value of money in that lead counsel has not been paid contemporaneously with their work in this case. See Vizcaino v. Microsoft Corp., 290 F.3d 1043, 1051 (9th Cir.2002) (citing Gates v. Deukmejian, 987 F.2d 1392, 1406 (9th Cir.1992)) ("Calculating fees at prevailing rates to compensate for delay in receipt of payment was within the district court's discretion."); Fischel v. Equitable Life Assurance Society, 307 F.3d 997 (9th Cir.2002) (citing In re Coordinated Pretrial Proceedings in Petroleum Products Antitrust Litigation, 109 F.3d 602, 609 (9th Cir.1997)) (district court has discretion to compensate for delayed payment by either applying current rates or applying historical rates with a prime rate enhancement). Of course, the annual rise in an attorney's billing rate represents not only inflation but also the increased experience of the attorney, if different hourly rates are used (as the court does here) for lawyers of different experience and billing rates. But the inflationary effect should not usually grossly affect the lodestar calculation, unless the litigation is greatly prolonged, and such is not the case here. (The problem of "experience inflation" can be avoided altogether by the use of a "blended" hourly rate, which as explained presently has other advantages.) At any rate, the relevant data are these:

--------------------------------------------------------------------------
2005 Billing     Total        Total
Attorney        Experience      Rate (per hr)    Hours        lodestar
--------------------------------------------------------------------------
Sidener         20 years        $510             1027.50      $524,025.00
--------------------------------------------------------------------------
Bennett         31 years        $550               72.75      $ 40,012.50
--------------------------------------------------------------------------
Cera            23 years        $525                5.00      $  2,625.00
--------------------------------------------------------------------------
Barton          8 years         $360             $397.00      $144,990.00
--------------------------------------------------------------------------
Giblin          9 years         $340               66.50      $ 22,610.00
--------------------------------------------------------------------------
Dirksen         4 years         $300              210.50      $ 63,150.00
--------------------------------------------------------------------------
Kim             2 years         $200               48.25      $  9,650.00
--------------------------------------------------------------------------
Investigator    n/a             $230              401.00      $ 92,230.00
--------------------------------------------------------------------------
Paralegals      n/a             $163               29.50      $  4,805.00
--------------------------------------------------------------------------
Totals     2258.00      $904,097.50
-------------------------

The above figures are as reported in the Sidener declaration (Doc. # 167), with the exception that the court has aggregated all paralegal hours and converted individual paralegal rates to a blended rate.

The Sidener declaration breaks out the hours expended by lead counsel into five categories: (1) investigation; (2) discovery; (3) pleadings, briefs and motions; (4) court appearances; and (5) settlement. This is an especially helpful compromise between reporting hours in the aggregate (which is easy to review, but lacks informative detail) and generating a complete line-by-line billing report (which offers great detail, but tends to obscure the forest for the trees). Upon review of this breakout, the court concludes that counsel spent a reasonable number of hours on this litigation. Additionally, two aspects are notable: First, the bulk of the hours (more than 75%) were spent on investigation and settlement. Second, about 29% of the hours expended (and about 37% of the fees accumulated) were devoted by Steven Sidener, the lead partner on this case, to settlement. This captures the importance to this case of settlement preparations and discussions.

The unusually large fraction of senior attorney time devoted to settlement is also relevant to counsel's hourly rate. Billing all attorney time at a "blended" hourly rate is probably appropriate in most lodestar cross-check computations for two reasons. First, it simplifies calculations. Second, it encourages economically efficient allocation of work within class counsel's firm. Third, it avoids the problem of "experience inflation" mentioned above. But the central role of settlement negotiations in this litigation — and the central role of attorney Sidener in those negotiations — suggest that typical blended hourly rates (perhaps in the $200 per hour neighborhood) are inappropriate here. Cf. Allen v. Bay Area Rapid Transit District, 2003 WL 23333580 at *7 (N.D.Cal.) (concluding in a civil rights case on which two attorneys worked that "the court has no basis for employing a reasonable hourly billing rate in its lodestar calculation either greater or lower than the approximate local average of $150 discernible from publicly available data"); Albion Pacific Property Resources, LLC v. Seligman, 329 F Supp 2d 1163, 1178 (N.D.Cal.2004) (holding, on a successful motion to remand, "plaintiff is entitled to an above-average hourly rate if it can demonstrate that its counsel were more efficient than reasonably competent counsel would have been" but that "[p]laintiff has made no such showing" and accordingly awarding "a reasonable attorney fee at an hourly rate of $190/hr for attorneys"); Yahoo!, Inc. v. Net Games, Inc., 329 F Supp 2d 1179, 1185, 1188 (N.D.Cal.2004) (explaining that "the average market rate in the local legal community as a whole is a better approximation of the hourly rate that would be charged by reasonably competent counsel than the actual billing rate charged by a single attorney" and that "[f]ocusing on the requested number of hours as a whole rather than the requested hourly rates of individual attorneys produces better attorney fee decisions").

The court understands that such blended rates typically depend on the overall billing mix including substantial time expended in discovery by junior attorneys with relatively low billings rates. Here, little time (3.8% of the total hours) was spent on discovery, but a large fraction of lead counsel's overall time was spent — and necessarily so, it seems — by a senior attorney in settlement discussions and preparation. Accordingly, the court concludes that use of a blended hourly rate is a poor fit for this case.

But the court must find some objective source for setting counsel's hourly rates; the court cannot simply look at a lone out-of-context dollar figure and pronounce it "reasonable." Because the court has rejected the use of a blended rate here, another problem arises: The court will need a variety of rates to account for the various attorneys' different levels of experience. One well-established objective source for rates that vary by experience is the Laffey matrix used in the District of Columbia. See http://www.usdoj. gov/us ao/dc/Divisions /Civil_Division /Laffey_Matrix _4.html (citing Laffey v. Northwest Airlines, Inc., 572 F. Supp. 354 (D.D.C.1983), aff'd in part, rev'd in part on other grounds, 746 F.2d 4 (D.C.Cir.1984)).

Under the 2005 Laffey matrix, attorneys with 20 or more years of experience bill $390/hour; attorneys with 8-10 years of experience bill $280/hour; attorneys with 4-7 years of experience bill $225/hour; attorneys with 3 or fewer years of experience bill $185/hour; and paralegals bill $110/hour. These figures are, however, tailored for the District of Columbia, which has a somewhat lower cost of living than the San Francisco Bay area (in which lead counsel's firm operates); the court will adjust these figures accordingly. The locality pay differentials within the federal courts — which, like law firms, employ lawyers and legal support staff — can approximate this difference. See http://jnet.ao.dcn/ Human_Resources /Pay_Tables/2 005_Pay _Tables/ Judiciary_ Salary_Plan _Pay_ Tables_ 2005.html. The Washington-Baltimore area has a +15.98% locality pay differential; the San Francisco-Oakland-San Jose area has a +26.39% locality pay differential. Thus, adjusting the Laffey matrix figures upward by approximately 9% will yield rates appropriate for the Bay area.^[1]

Applying this adjustment and rounding, the court obtains the following rates: Attorneys with 20 or more years of experience bill $425/hour; attorneys with 8-10 years of experience bill $305/hour; attorneys with 4-7 years of experience bill $245/hour; attorneys with 3 or fewer years of experience bill $200/hour; and paralegals bill $120/hour. Not having a suitable substitute for the investigator's billing rate, the court will accept the value of $230/hour from the Sidener declaration. Reproducing the table above, but substituting these values and recomputing the totals yields:

--------------------------------------------------------------------------
2005 Billing     Total       Total
Attorney          Experience     Rate (per hr)    Hours       lodestar
--------------------------------------------------------------------------
Sidener           20 years       $425             1027.50     $436,687.50
--------------------------------------------------------------------------
Bennett           31 years       $425               72.75     $ 30,918.75
--------------------------------------------------------------------------
Cera              23 years       $425                5.00     $  2,125.00
--------------------------------------------------------------------------
Barton            8 years        $305              397.00     $121,085.00
--------------------------------------------------------------------------
Giblin            9 years        $305               66.50     $ 20,282.50
--------------------------------------------------------------------------
Dirksen           4 years        $245              210.50     $ 51,572.50
--------------------------------------------------------------------------
Kim               2 years        $200               48.25     $  9,650.00
--------------------------------------------------------------------------
Investigator      n/a            $230              401.00     $ 92,230.00
--------------------------------------------------------------------------
Paralegals        n/a            $120               29.50     $  3,540.00
--------------------------------------------------------------------------
Totals    2258.00     $768,091.25
------------------------

In the end, this substitution reduces lead counsel's lodestar fee by about 15% from the figure claimed in the Sidener declaration (Doc. # 167).

2

The court now turns to the lodestar cross-check, which entails evaluation of the multiplier implied by lead counsel's requested fee (15% percent of a $18.4 million settlement fund, or $2.76 million) and lead counsel's lodestar fee (computed above as $768,091.25). These data imply a multiplier of 3.59. Is a multiplier of 3.59 reasonable in this case?

Unfortunately, there are not (to the court's knowledge) any reliable compilations of the multipliers awarded in common fund class actions against which the court could quantitatively compare this 3.59 multiplier. Multiplier values are reported in the Class Action Reports data set upon which the court has relied in previous orders. See Stuart J Logan, Jack Moshman & Beverly C Moore, Attorney Fee Awards in Common Fund Class Actions, 24 Class Action Rep 167 (2003). But the Class Action Reports data are contaminated because the courts making the reported awards have been inconsistent in reporting a base hourly rate. Floating without an empirical tether, the court shares Judge Shadur's sentiment in Comdisco that "lodestar multipliers * * * seem most often to represent an effort to provide some non-existent garb for the proverbial emperor who has no clothes. * * * Unfortunately, any ruling that a multiplier (say) of 2.5 is reasonable, while a multiplier (say) of 3.5 or 4 is not, too often makes it possible to paper over a result-driven conclusion with a purported semblance of precision" 150 F Supp 2d at 948 n. 10.

The court has an idea for a solution (albeit one that is impossible in this case): Class counsel's risk of nonrecovery (or the obverse, their probability of success) is central to the multiplier. See In re Washington Public Power Supply System Securities Litigation, 19 F.3d 1291, 1301 (9th Cir.1994) (discussing risk multipliers). Therefore, pursuant to FRCP 23(g)(1)(C)(iii) and 23(h)(3), class counsel ought to be required at the outset of the case to file (ex parte and under seal) a candid quantitative assessment of their probability of success and the likely effort entailed in obtaining various recoveries. The information submitted to the court should address not only the probability of total nonrecovery, but also the relative likelihood of various levels of recovery, for it is only against the full continuum of possible results that the court can assess the true success that class counsel achieves (or fails to achieve). Moreover, counsel should provide their assessments of the likely time frame of the litigation and the costs they expect to incur. Indeed, this information should capture not only the "risk of nonrecovery" aspect of a multiplier, see id, but also whether counsel achieved exceptional results for the class, see, e.g., Van Gerwen v. Guarantee Mut. Life Co., 214 F.3d 1041, 1045 (9th Cir.2000) (discussing multipliers in fee-shifting context). To avoid any concern that such evaluation would contaminate the judge's decisions in the case (or reveal confidences to opposing counsel), the evaluation would remain under seal until needed to assess an application for fees and expenses. A form order requesting this information and setting conditions for its submission is annexed to this order.

It is entirely reasonable to expect class counsel to make such assessments. Any economically rational individual plaintiff retaining counsel would surely ask his attorneys for just this information: "Honestly, what do I stand to recover, and how much is it going to cost me in fees?" A sensible lead plaintiff executing the ideal model of the PSLRA's "most adequate plaintiff" would obtain exactly this kind of information. Should not the court also seek it, especially if there is any uncertainty that lead plaintiff has searched for counsel and negotiated at arms' length for a reasonable fee? Further, of course, "entrepreneurial plaintiff's attorneys" (Professor Coffee's phrase) would perform just such an analysis before commencing the long and arduous effort that often characterizes class litigation. See John C Coffee, Jr, Understanding the Plaintiff's Attorney: The Implications of Economic Theory for Private Enforcement of Law Through Class and Derivative Actions, 86 Colum L Rev 669 (1986).

Of course, in the class action context, knowing that the court would use such an ex ante assessment to award a fee (should counsel obtain a recovery), this exercise will no doubt tempt counsel to emphasize the difficulty of the task ahead of them. The court needs to recognize this danger. But even with it, bearing the risk of an unduly pessimistic evaluation of the prospects at the outset of a case or an overblown assessment of the effort required seems no worse (and may be better) than facing inflated claims of accomplishment when class counsel seek approval of a compromise and an award of fees at the end of a case. An ex ante assessment has to make economic sense or it loses credibility. Counsel who claim a one in twenty chance of a $10 million recovery and estimate 10,000 hours of attorney time to obtain it are obviously blowing smoke. There is a limit on feigned pessimism, for counsel's prediction has to square with their own economic rationality: Predictions that are too pessimistic raise the question why counsel got involved in the case in the first place. Thus an analysis using a blended hourly rate or the Laffey matrix could reveal whether counsel's ex ante assessment makes sense as a matter of law firm economics — and therefore is credible.

Furthermore, the court can evaluate counsel's ex ante assessment by the same standards that apply to many other aspects of a court's work. On the one hand, an evaluation that is not substantiated by reasoned analysis and comparable cases would be entitled to little weight. On the other hand, an evaluation that laid out in detail the factors that went into the evaluation and presented data to back it up would carry a good deal of weight even if the predicted outcome, length of proceedings and costs differed from what actually occurred.

Even absent a true ex ante assessment, the parties ought to be able to compile — at least if the case is settled, as often happens in successful common fund cases — a record for the court that demonstrates their assessments of their legal positions, their predictions of liability exposure, their settlement positions and negotiations and so on. Perhaps a complete record is too much to ask for, but surely in every case some contemporaneous evidence exists that is probative of the parties' assessments of their likelihood of success. Because the court has not asked for this information and did not ask for an ex ante assessment in this case, the court here proceeds from a rather more limited record.

Based on this limited record, the implied multiplier of 3.59 gives the court considerable pause. It exceeds the multipliers the court has in its experience encountered and observed in other common fund securities class actions. The court can investigate this intuition further by positing a pair of scenarios — one optimistic, one pessimistic — of the sort that plaintiffs' counsel almost certainly considered at the inception of this litigation. Each scenario has several individual components representing the probability of recovering a certain amount against a certain defendant or group of defendants; using standard probability, the expectation value of the total scenario can be computed. Here are two scenarios that seem possible to the court in light of its experience with company defendants, accountant defendants and insurance coverage layers in securities class actions in the context of the insurance coverage and resources of the defendants in this litigation:

Scenario 1: Optimistic Scenario
-----------------------------------------------------------
100% chance of recovering $3M from insurance       $3.00M
carriers
-----------------------------------------------------------
66.7% chance of recovering $4M more from           $2.67M
company defs
-----------------------------------------------------------
50% chance of recovering $10M from                 $5.00M
accountant
-----------------------------------------------------------
Total expectation: $10.67M
--------
Multiplier based on $18.4M settlement:   1.72
--------
Scenario 2: Pessimistic Scenario
-----------------------------------------------------------
50% chance of no recovery at all                   $0.00M
-----------------------------------------------------------
50% chance of recovering $7M from company          $3.50M
defendants
-----------------------------------------------------------
25% chance of recovering $10M from                 $2.50M
accountant
-----------------------------------------------------------
Total expectation: $6.00M
---------

Multiplier based on $18.4M settlement:              3.07
--------

These scenarios suggest that a multiplier between 1.72 and 3.07 is reasonable.

In fairness, there is one reason to think that a multiplier toward the higher end of this wide range is appropriate here: Lead counsel have generated an unusually high recovery per hour invested. Even valuing the HPL stock at $0.20 per share, the settlement consideration is worth $18.4 million — or about $8,148 per hour invested by lead counsel. If the court would approve a 15% fee award, the net recovery per hour would be about $6,926. Looking to the data compiled in Logan et al, 24 Class Action Rep 167, the court notes that in class actions with common funds of $10 to $20 million, the average net recovery per hour is $2,943, with a range of $292 to $31,996. Of the 63 cases in this range reported in Logan et al, this case would rank 6th highest in recovery per hour. In fact, this recovery per hour even exceeds the median recovery per hour of $4,946 seen in "megafund" cases (those with common funds that exceed $100 million, and thus should exhibit the greatest economies of scale).

On the one hand, these impressive results may signal lead counsel's unusual skill and efficiency. On the other hand, they may signal a case ripe for generous settlement. But the latter possibility seems less likely: As the court has observed in its prior orders, establishing liability against — and hence extracting a settlement from — defendants other than Y David Lepejian (whose participation in the fraud here was demonstrably knowing) was far from certain, and it is clear that settlement negotiations in this case were anything but a cakewalk.

Nonetheless, the court has difficulty squaring a multiplier of 3.59 with what it knows about this litigation. The lodestar cross-check has thus signaled that the 15% fee proposed by lead counsel with lead plaintiff's support leads to a multiplier higher than appears reasonable. The court concludes, in the exercise of its discretion in light of the foregoing discussion, that a reasonable fee award is somewhat lower than the requested 15%.

The amount of reduction is the next question. For reference, the court has tabulated various fee percentages, the resulting fee and the corresponding multiplier (valuing the settlement at $18.4 million):

Fee           Implied
Percentage    (millions)      Multiplier
15%          $2.76           3.59
14%          $2.58           3.35
13%          $2.39           3.11
12%          $2.21           2.87
11%          $2.02           2.63
10%          $1.84           2.39
9%          $1.66           2.16
8%          $1.47           1.91
7%          $1.29           1.67

Based on the court's conclusions above about reasonable multipliers in this case, it appears that a reasonable percentage fee in this case ranges from about 8% to about 12%. Because (1) lead plaintiff supports a 15% fee award for lead counsel and (2) the recovery in this case is exceptionally high for the amount of attorney time invested, the court tends to the high side of this range. In its discretion the court fixes lead counsel's fee award at 11% of the common fund.

Accordingly, the court awards lead counsel $1,870,000 cash and 770,000 shares of HPL stock from the common fund. With respect to the award of HPL stock, to the extent that such an award represents securities that are not registered pursuant to the Securities Act of 1933, the court finds, pursuant to section 3(a)(10) of the Securities Act of 1933 that under the circumstances, the exchange of these shares for a portion of lead counsel's services in this action is fair.

IV

This order treads on somewhat unfamiliar ground, yet the court believes no less is necessary to discharge its obligation under FRCP 23(h) to award a reasonable fee. The court's excursion has led it to take up legal issues that have not been briefed by counsel, and this order rests in some measure on a factual record that counsel might have supplemented, had they known the court's view of the law. In other words, lead counsel have perhaps not been offered a full opportunity to be heard on the issues addressed herein — an opportunity that this court must now in fairness afford them, particularly given the deferential standard of appellate review applied to fee awards under Rule 23(h).

Accordingly, the clerk shall enter final judgment on the award of fees and expenses as described herein, but the court expressly grants lead counsel leave to file a motion and supporting papers pursuant to FRCP 60(b)(6) for relief from that judgment. Such a motion for relief from judgment shall be filed within the time allotted by FRAP 4(a)(1)(A) — that is, if lead counsel wish to challenge this order here or in the Court of Appeals, they must act within the time allotted by FRAP 4(a)(1)(A); otherwise, distribution of the settlement fund may commence. Along with any motion for reconsideration, lead counsel should advise the court whether an interim distribution of the uncontested portion of the settlement fund is warranted.

V

In sum, the court GRANTS lead plaintiff's motion for an award of attorneys' fees and costs (Doc. # 142). Out of the common fund, the court awards lead counsel expenses of $59,434.57 and fees of $1,870,000 cash plus 770,000 shares of HPL stock. These amounts shall bear interest at the same rate and from the same date as the settlement funds.

IT IS SO ORDERED.

APPENDIX: SAMPLE ORDER

ORDER

Pursuant to FRCP 23(g)(1)(C)(iii) and 23(h)(3), the court ORDERS putative class counsel to submit, under seal, not later than __________ counsel's good faith quantitative assessments of:

• their probability of prevailing on plaintiffs' claims, including their likelihood of achieving various levels of recovery;

• the hours of attorney and legal support staff time they reasonably anticipate spending to reach certain stages in the litigation and to achieve various levels of recovery; and

• the reimbursable costs they expect to incur to reach certain stages in the litigation and to achieve various levels of recovery;

stating with particularity the reasons for these assessments. See In re HPL Technologies Securities Litigation, 366 F. Supp. 2d 912, 2005 WL 941586 (N.D.Cal 2005).

Counsel shall contact the court room deputy for direction in making this submission, which shall be made in paper form, ex parte and under seal. The clerk is directed not to unseal this submission or permit its inspection by anyone except upon notice to the parties and further order of the court.

IT IS SO ORDERED.

NOTES

[1] (126.39 — 115.98)/115.98 = 0.08976, or about 9%.