1 2 3 4 5 6 UNITED STATES DISTRICT COURT 7

NORTHERN DISTRICT OF CALIFORNIA

8

9

10 POLICE RETIREMENT SYSTEM OF ST. LOUIS, on behalf of itself and similarly- 11 situated plaintiffs, No. C 19-04744 WHA

12 Plaintiff,

13 v.

ORDER RE MOTIONS FOR FINAL

APPROVAL OF CLASS ACTION

14 GRANITE CONSTRUCTION SETTLMENT AND ATTORNEY’S

INCORPORATED, JAMES H. ROBERTS, FEES AND MOTION TO SEAL

15 JIGISHA DESAI, and LAUREL J.

KRZEMINSKI,

16 Defendants. 17

18 INTRODUCTION 19 In this securities class action, lead plaintiff moves for final approval of the settlement. 20 To the extent stated, the motion is GRANTED. Class counsel and intervenor counsel’s motions 21 for attorney’s fees and expenses are GRANTED IN PART AND DENIED IN PART. Intervenor 22 counsel’s motion to seal is DENIED. 23 STATEMENT 24 Prior orders have summarized the facts in detail (Dkt. Nos. 98, 263). In short, after 25 defendant Granite Construction Incorporated issued a restatement, the parties settled. Court- 26 appointed lead plaintiff Police Retirement System of St. Louis filed a motion for preliminary 27 approval, along with a stipulation of settlement in April 2021 (Dkt. No. 176). That agreement 1 provided for $129 million in a class fund. Before preliminary approval, a class member, Arash 2 Nasseri, objected to the plan of allocation, moved to intervene and for appointment as co-lead 3 plaintiff, all on the supposed ground that the plan of allocation should have allocated more to 4 claimants who had Section 11 claims as well as Section 10(b) claims (as opposed to those with 5 only Section 10(b) claims). The plan of allocation indeed did award them more, but Nasseri 6 contended even more should be awarded (Dkt. No. 179-1 at 10–16). Nasseri also served as 7 named plaintiff for a putative uncertified class of Section 11 claimants suing defendants in 8 California superior court. His lawyer, Jason Forge of Robbins Geller Rudman & Dowd LLP 9 moved for appointment as co-lead counsel. Intervention was granted, and the remaining 10 motions denied (Dkt. No. 232). 11 All sides agreed to return to mediation. Nasseri, Retirement System, and defendants 12 agreed to increase the portion of the settlement going to the Section 11 claimants. With the aid 13 of Judge Joseph C. Spero, they landed on a multiplier of 2.2, meaning that those class members 14 with Section 11 claims as well as Section 10(b) claims would receive 2.2 times as much 15 recovery as those with only Section 10(b) claims (Dkt. Nos. 194; 220 at 3–4). The Court 16 appointed as special master Bruce Ericson, a trial lawyer expert in securities litigation. He 17 recommended a multiplier of 2.21. Our prior order herein dated October 6, 2021, granted 18 preliminary approval and adopted a 2.21 multiplier (Dkt. Nos. 232, 258, 263). 19 Since that order, claims administrator Epiq Class Action and Claims Solutions Inc. sent 20 notice to the class. The Court erroneously received fourteen notices intended for class 21 members. Prior orders informed the parties and made all but one of those notices available for 22 inspection in late December 2021. Scheduling difficulties caused the final notice to be absent 23 from chambers at the time of the inspection. Attorney Peter Borkon of Bleichmar Fonti & 24 Auld LLP (BFA), representing Retirement System, accepted the offer to inspect the notices 25 and filed two statements in response. The second also appended a declaration by Jessie Mahn, 26 Epiq’s project manager. The declarations detailed remedial measures to resend notices, with 27 the cost of supplemental notices to be borne by counsel, not the class. Epiq now certifies that it 1 certification, lead plaintiff offered evidence that there were between 40 and 49 million in 2 Granite common shares outstanding on the New York Stock Exchange (NYSE) during the 3 class period. An additional 453 institutional investors held Granite common stock. Weekly 4 trading volume averaged 2.16 million during the class period. As of our hearing on February 5 24, an unknown number of all affected shares had filed claims. Class counsel were ordered to 6 provide a report regarding the percentage of shares claimed to date, as an estimated percentage 7 of total shares (Dkt. Nos. 120-1 at ¶¶ 25, 28, 65, 71; 263; 270–73; 278; 286 at 3; 298). 8 In December 2021, class and intervenor counsel moved for attorney’s fees. Class 9 counsel responded to intervenor counsel’s motion, while intervenor counsel “objected” to class 10 counsel’s motion. Both replied in support of their own motions. Intervenor counsel filed a 11 sur-reply (Dkt. Nos. 274, 276, 280, 282–85). In February 2022, Retirement System moved, 12 unopposed, for final approval (Dkt. No. 286). 13 ANALYSIS 14 “The class action device, while capable of the fair and efficient adjudication of a large 15 number of claims, is also susceptible to abuse and carries with it certain inherent structural 16 risks.” Officers for Just. v. Civ. Serv. Comm’n of City & Cty. of San Francisco, 688 F.2d 615, 17 623 (9th Cir. 1982). A settlement purporting to bind absent class members must be fair, 18 reasonable, and adequate. See FRCP 23(e). Rule 23(e)(2) requires district courts to employ a 19 two-step process: First, the parties must show the district court will likely be able to approve 20 the proposed settlement. Second, the district court must hold a hearing to make a final 21 determination of whether the settlement is fair, reasonable, and adequate. We have arrived at 22 step two. 23 Our court of appeals recently explained that the final fairness assessment must analyze 24 the eight Churchill factors: (1) the strength of the plaintiff’s case; (2) the suit’s risk, expense, 25 complexity, and the likely duration of further litigation; (3) the risk of maintaining class-action 26 status throughout the trial; (4) the amount offered in settlement; (5) the extent of discovery and 27 the stage of the proceedings; (6) the experience and views of counsel; (7) the views of a 1 settlement. See Kim v. Allison, 8 F.4th 1170, 1178–79 (9th Cir. 2021) (quoting In re 2 Bluetooth Headset Prods. Liab. Litig., 654 F.3d 935, 946 (9th Cir. 2011)); Churchill Vill., 3 L.L.C. v. Gen. Elec., 361 F.3d 566 (9th Cir. 2004). Additionally, Rule 23(e)(2) requires the 4 district court to consider an overlapping set of factors. These include the adequacy of the 5 notice procedure and “the terms of any proposed award of attorney’s fees.” Kim, 8 F.4th at 6 1179. These also require the district court to scrutinize the settlement for evidence of collusion 7 or conflicts of interest, and to review other, relevant factors before deeming the settlement fair. 8 See Briseño v. Henderson, 998 F.3d 1014, 1023–26 (9th Cir. 2021). Our notice regarding 9 factors to be evaluated for proposed class settlement, filed herein November 21, 2019, also lists 10 considerations that this order will assess (Dkt. No. 50). 11 In short, in exchange for the dismissal of this action with prejudice and a release of all 12 claims, defendants agree to pay $129 million to be allocated to class members on a pro rata 13 basis. Our plan of allocation contemplates that Section 11 claimants will recoup 2.21 times the 14 amount that class members who only assert Section 10(b) claims will recoup (see Dkt. Nos. 15 258, 263). Just one class member has opted out and no one has objected to the settlement or 16 plan of allocation (Mahn Decl. ¶ 9; Borkon Decl. ¶ 103). 17 1. THE CHURCHILL FACTORS. 18 We begin with the eight Churchill factors. They support settlement. 19 First, the strength of the claims for both Section 10(b) and Section 11 class members 20 favors this settlement. Though as Special Master Ericson detailed, all claims ranked as “above 21 average,” substantial risks remain (Dkt. No. 258 at 9). On one hand, at trial, all plaintiffs 22 would come armed with Granite’s “big R” restatement, which revised relevant financials for 23 the entire class period of February 17, 2017, to October 24, 2019. As an “error” restatement, it 24 fell short of the strongest form of “irregularity” restatement (id. at 4, 7; Dkt. No. 159). 25 Evidence of loss causation could flow from Granite’s corrective disclosures in July, August, 26 and October of 2019, at which time stock prices fell roughly 40%. The 1934 Act claimants, 27 however, might have struggled to show loss causation prior to July 2019, leading Special 1 average (Dkt. No. 258 at 3). Granite no longer stands by its financials, and this repudiation 2 might have facilitated proving scienter at trial. Several problems remained, however. For 3 instance, it could have been hard to show, important for the Section 10(b) claimants, scienter as 4 to the earlier alleged misstatements. Defendants could have painted the forward-looking 5 disclosures (for instance, revelations by Granite that its losses looked comparably better on a 6 pro rata basis than losses recognized by the joint ventures throughout 2018) as poor business 7 judgments and as lacking intent to defraud (Amd. Comp. Exh. B, Dkt. No. 69-2). 8 Second, the risk, expense, complexity, and likely duration of the case supported 9 settlement. As explained, plaintiffs faced notable risks. Substantial expense would have 10 accompanied trial planning, including preparing any or all of the eight confidential witnesses 11 from the complaint. Continuation of discovery and multiple legal claims all made the case 12 likely longer and more complex. 13 Third, the risk of maintaining class action status throughout trial would not have posed a 14 serious threat. The class was certified in January 2021 and classes are regularly maintained in 15 securities suits (Dkt. No. 127). This factor rates as neutral. 16 Fourth, the amount of the award supports the settlement agreement. After all, even “a 17 cash settlement amounting to only a fraction of the potential recovery does not per se render 18 the settlement inadequate or unfair.” In re Mego Fin. Corp. Sec. Litig., 213 F.3d 454, 459 (9th 19 Cir. 2000), as amended (June 19, 2000)) (quoting Officers for Just. v. Civ. Serv. Comm., 688 20 F.2d 615, 628 (9th Cir. 1982) (approving securities settlement of one-sixth recoverable 21 damages). Class counsel offer that this settlement amounts to 20% to 30% of the class’s 22 estimated range of recoverable damages ($424–$670 million) (Borkon Decl. ¶ 101). Class 23 counsel cite the Cornerstone Research Group for the proposition that a settlement of $129 24 million would fall in the top four in terms of absolute dollar recovery for a securities class 25 action in our district in about a decade (id. ¶ 100). (Another securities class action in our 26 district has litigation around settlement pending, which settlement, if granted at final approval, 27 would put ours in the top five instead. See In re Twitter Inc. Sec. Litig., No. C 16-5314 (N.D. 1 Class counsel also cite research that found a median settlement of 5.4% estimated 2 recovery for suits with claims under Sections 10(b) and 11 (Dkt. No. 176 at 13). See 3 Cornerstone Research, Securities Class Action Settlements, 2020 Review and Analysis, at 7, 4 https://www.cornerstone.com/wp-content/uploads/2021/12/Securities-Class-Action- 5 Settlements-2020-Review-and-Analysis.pdf. It bears emphasis now and later in this order that 6 the $129 million settlement was achieved by class counsel (Peter Borkon and his firm) and not 7 by intervenor counsel. 8 Fifth, the stage of the proceedings supports settlement. Class counsel investigated this 9 case in great depth. This effort culminated in the consolidated amended complaint. Lead 10 plaintiff and defendants litigated a complex motion to dismiss. When settlement finally 11 arrived, lead plaintiff and defendants had engaged in extensive discovery. Class counsel had 12 reviewed two million pages of discovery, taken three depositions, and prepared for thirteen 13 more. The parties resolved various discovery disputes and litigated one. True, defendants did 14 not oppose class certification, but class counsel had also moved for partial summary judgment 15 before settlement (Dkt. No. 159). By that point, discovery and litigation had given each party a 16 fair sense of the other side’s strengths and weaknesses. 17 Sixth, the abilities and views of counsel support settlement. Class counsel have extensive 18 experience in securities class actions and have pursued cooking-the-books theories of PLSRA 19 violations in at least one other case (see, e.g., Dkt. No. 277-7, Borkon Decl. Exh. G at 9, 23– 20 24). This order finds defense counsel and intervenor counsel also experienced. 21 Seventh, there was no governmental participant in this action. This factor is neutral. 22 Eighth, no class member has objected to the proposed settlement and only one has opted 23 out. This factor moderately supports settlement (Mahn Decl. ¶ 7, Dkt. No. 281-1). 24 In sum, the Churchill factors support final approval of the $129 million class settlement. 25 2. THE PLAN OF ALLOCATION. 26 This order agrees that claimants who have both Section 11 and Section 10(b) claims 27 should receive more on a pro rata basis than those with only Section 10(b) claims, mainly 1 proposed to allocate more to those with Section 11 claims. The original multiplier was 1.6. 2 After the contention by Nasseri, the multiplier wound up at 2.21. Based on our special 3 master’s report, this order finds that 2.21 is a fair and reasonable multiplier. 4 Again, it bears emphasis that the adjustments of the multiplier came after class counsel 5 had already achieved a global settlement of $129 million. Nasseri and his counsel were not 6 involved in litigating this case or in negotiating that $129 million success. 7 3. RULE 23 AND THIS COURT’S SETTLEMENT FACTORS. 8 As stated, consideration of the Churchill factors alone is not enough. Kim, 8 F.4th at 9 1179. This order accordingly evaluates fairness, reasonableness, and adequacy under Rule 10 23(e)(2) and our factors to be evaluated for a proposed class settlement, to the extent not 11 covered above (Dkt. No. 50). 12 First, the claim procedure for class members was appropriate. Each class member was to 13 receive a claim form with instructions to detail all purchases or acquisitions and sales between 14 February 17, 2017, and January 22, 2020, of Granite common stock; all Granite common stock 15 held at the opening and the close of trading on February 17 and on January 22; and shares 16 obtained in the merger (Dkt. No. 176-1 Exh. A-1 at 22, n. 3). If a claimant had obtained shares 17 in the merger, then the claimant was to be asked to detail sales of Granite common stock up 18 through the specified date on which the notice was issued. In order to recover, each claimant 19 was required to execute and return the claim form. The notice, moreover, was straightforward 20 and comprehensible. There was no burdensome opt-out procedure. Epiq had apparently, as of 21 February 28, 2022, timely sent over 57,000 notices (some via nominees), received 28,114 22 claims, and processed approximately 23,000 claims. Both Epiq and class counsel affirmed 23 multiple times that Epiq would continue to accept and process claims received after the 24 January deadline, until such time as doing so would impact disbursement (Mahn Decl. ¶¶ 6–7, 25 Dkt. No. 290; Dkt. No. 262; Mahn Decl. ¶¶ 2–3, Dkt. No. 281-1; Dkt. No. 298 at 1). 26 At the final approval hearing, the Court requested certain follow up information 27 concerning claim processing and has now received and read all follow up responses. These 1 the same day as our final approval hearing), and from Attorney Forge (Dkt. Nos. 298–302). 2 The Court is convinced that there is no need for further action and that approval of the final 3 settlement is in order. Except to the extent otherwise stated in this order, any request for relief 4 contained in the post-hearing letters to the Court is DENIED. 5 Second, the release of claims against defendants bears a permissible scope. The 6 settlement releases defendants from claims related to the specific allegations of the 7 consolidated amended complaint and to claims arising from or relating to the purchase of 8 Granite stock allegedly inflated in price during the class period. It expressly omits releases of 9 any claims related to a separate shareholder derivative suit (English v. Roberts, et al., No. C 10 20-03116 (N.D. Cal.)). Release of intervenor plaintiff’s state claims also passes the test from 11 our court of appeals. The facts arise out of “an identical factual predicate.” Hesse v. Sprint 12 Corp., 598 F.3d 581, 591 (9th Cir. 2010). The parties have explicitly limited the release to a 13 permissible scope (Dkt. No. 176-1 ¶ 1.25). 14 Third, the remaining factors for consideration all support settlement. Parties left the 15 request for attorney’s fees to the judge’s discretion in settlement. Relatedly, the terms and 16 timing of attorney’s fees, as discussed in detail below, are permissible. This order finds, as 17 further considered below, class counsel and Retirement System have adequately represented 18 the class. No reversion provision appears in the settlement agreement. This order finds no 19 evidence of collusion or conflicts of interest. As noted, key phases of settlement benefitted 20 from the input of Judge Spero. Additional scrutiny by Special Master Ericson increased our 21 confidence that the plan of allocation is fair. Lastly, Retirement System has not requested an 22 incentive payment and stands to benefit only the same as all class members. Intervenor 23 Nasseri has requested an incentive payment for himself, but this request will not erect a barrier 24 to final approval. On balance, these facts favor settlement. 25 In sum, this Court’s factors as well as the factors listed in Rule 23 support settlement. 26 Final approval of the settlement and plan of allocation is GRANTED. 27 4. ATTORNEY’S FEES AND COSTS. 1 Class counsel propose a percentage fee of 16.33% with a “reasonable” portion of the 2 award going to intervenor counsel. Intervenor counsel propose a 12% fee award, with 29% of 3 that figure to intervenor counsel and the balance paid to class counsel. 4 A district court must ensure that attorney’s fees are “fair, adequate, and reasonable.” 5 Staton v. Boeing Co., 327 F.3d 938, 963–64 (9th Cir. 2003). Where settlement creates a large 6 fund “the district court has discretion to use either a percentage or lodestar method.” Ibid. Our 7 court of appeals has recognized twenty-five percent of the common fund as a “benchmark 8 award for attorney’s fees.” Hanlon v. Chrysler Corp., 150 F.3d 1011, 1029 (9th Cir. 1998). 9 The district court must always “close[ly] scrutin[ize]” the request. In re Optical Disk Drive 10 Products Antitrust Litig., 959 F.3d 922, 930 (9th Cir. 2020). Large-fund settlements like ours 11 warrant exacting scrutiny. When awarding twenty-five percent of a very large fund “would 12 yield windfall profits for class counsel in light of the hours spent on the case, courts should 13 adjust the benchmark percentage or employ the lodestar method instead.” Bluetooth, 654 F.3d 14 at 942. This order finds the percentage method with lodestar cross-check appropriate under the 15 circumstances. Counsel agree on this much (Dkt. Nos. 274 at 10; 276 at 7). 16

A. PERCENTAGE FEE AWARDS.

17 Class counsel request $21,060,000, or approximately 16.33%, of the $129 million 18 settlement as a total fee award (Dkt. No. 284 at 1; Dkt. No. 176-1 ¶ 1.28). Intervenor counsel 19 request a different total fee, 12% ($15,548,000). The correct answers follow: we will first 20 deduct attorney’s expenses from the common fund. Then, we will award attorney’s fees of 21 16.33% of the remainder, or $20,921,494.16.1 22 To vary up or down from the twenty-five percent benchmark, a district court should 23 analyze (1) whether counsel “achieved exceptional results for the class”; (2) the risk to class 24 counsel; (3) whether counsel “generated benefits beyond the cash settlement fund”; (4) how 25 similar the proposed fee is to those in similar cases; and (5) the burden to counsel in 26

27 1 The total attorney’s fees award is calculated by subtracting from the class fund all expenses, 1 conducting the case. See Vizcaino v. Microsoft Corp., 290 F.3d 1043, 1048–52 (9th Cir. 2002) 2 (cleaned up); Hanlon, 150 F.3d at 1029. 3 As discussed above, (1), this settlement stands out for its size. Class counsel filed a 4 consolidated amended complaint with statements by eight whistleblowers. Within weeks, 5 defendants launched a related financial audit. Granite then restated relevant financial figures 6 for the entire class period. This work by class counsel rates high. Additionally, class counsel 7 negotiated during the court-ordered request-for-proposal (RFP) process and settled on 16.33%. 8 This deserves some deference. See In re Cavanaugh, 306 F.3d 726, 734, 736 (9th Cir. 2002) 9 (Dkt. No. 55).2 10 The risk on and burden to class counsel, (2) and (5), ranked as substantial during the 11 pendency of the case. Defendants pressed a motion to dismiss, the parties litigated numerous 12 discovery disputes, and plaintiff’s partial summary judgment motion faced considerable 13 challenges. When the settlement was announced, only plaintiffs had filed for summary 14 judgment and the dispositive motions deadline had not yet passed (see Dkt. Nos. 93, 176). The 15 burden likewise encompassed preparing for sixteen depositions and vast discovery review. 16 Overall, the risk favors class counsel’s requested fee award. 17 As for intervenor counsel, their work resulted in an increase in damages allocated to 18 Section 11 claimants but no change to the overall settlement. The risk was low and the burden 19 low, too. 20 Factor (3) includes the efforts of class counsel, which plausibly led to a restatement, 21 litigation of a motion to dismiss, and numerous discovery disputes. This factor favors their 22 requested fee award. For their part, intervenor counsel unquestionably achieved a benefit for 23 Section 11 claimants. 24 Finally, (4), a comparison to similar settlements favors class counsel’s proposal. 25 Between 2010 and 2021, attorney’s fee awards averaged 23.6% for funds between $100 and 26

27 2 The PLSRA requires lead plaintiff to solicit bids from counsel, which they must vet for cost, 1 $500 million. Janeen McIntosh and Svetlana Starykh, Recent Trends in Securities Class 2 Action Litigation: 2020 FULL-YEAR Review (Jan. 25, 2021), Figure 19, 3 https://www.lexology.com/library/detail.aspx?g=d4cb0c13-2a21-4c32-bb76-805f270d91ed. 4 This study also indicated that plaintiffs’ attorney’s fees and expenses represent a declining 5 percentage of settlement value as settlement size increases. Comparable settlements include: Case Settlement Attorney’s Noteworthy Facts 6 Amount Fees as % 7 of Common 8 Fund In re LendingClub $125 million 13.1% No restatement but quarterly report 9 Sec. Litig., 2018 WL “identified material weaknesses.” Also, CEO resigned based on 10 4 C5 a8 l.6 S6 e6 p9 t, . a 2t 4 * , 2 2 0(N 18.D );. c do en cefl ri tc ift i o caf ti in ot ne .r e st. Faced possible 11 see also C 16-02672 ECF No. 383 at 3. 12 In re Amgen Inc. Sec. $95 million 25% Nine-year litigation. High risk that Litig., 2016 WL primary evidence of scienter could 13 10571773, at *8 be excluded under FRE. (C.D. Cal. Oct. 25, 14 2016)(Judge Philip S. 15 Gutierrez). Luna v. Marvell Tech. $72.5 million 18.8% Risky litigation due to lack of 16 Grp., 2018 WL accounting restatement and denial from company auditor of any 1900150, at *3 (N.D. GAAP violation. Hotly-contested 17 Cal. Apr. 20, 2018). loss causation. 18 In re Diamond Foods, Ultimate value 14% Strong arguments that defendant Inc., Sec. Litig., No. of $123.8 violated GAAP and error 19 C 11-05386 WHA, million, made restatement was issued. 2014 WL 106826, at up of cash ($11 20 *4 (N.D. Cal. Jan. 10, million only) 21 2014). and stock whose value fluctuated 22 Our suit bears the closest resemblance to In re LendingClub. No officer resigned here 23 (compare In re Diamond Foods and In re LendingClub) and an error restatement was issued 24 (cf. Luna). But the accomplishment exceeded other suits’ because class counsel’s work 25 plausibly prompted our restatement. 26 By way of resolving objections, this order notes that class counsel protest intervenor 27 counsel’s sur-reply as filed without leave. Class counsel also argue waiver, citing intervenor 1 counsel’s failure to oppose (until the sur-reply) a 16.33% total fee award. The sur-reply 2 contravened our local rules, so this order has considered it only to explain disagreement with 3 the arguments therein. As for waiver, this order disagrees with intervenor counsel’s request for 4 12% and so, regardless, charts its own path. 5 Given the size of the fund, the efforts of counsel, and the relevant risks to the class, this 6 order finds class counsel’s proposal of a 16.33% fee award appropriate and a 12% award too 7 low. Once again, the 16.33% will be applied after attorney’s expenses have been subtracted 8 from the common fund. This is all subject to a lodestar cross-check. 9 B. DIVISION OF ATTORNEY’S FEE AWARD. 10 All agree that intervenor counsel deserve a portion of the fee. Here is the Court’s 11 assessment of what is fair and reasonable. 12 Intervenor counsel request 29% of the total fee award to match the percent of settlement 13 allocated to Section 11 claimants from the common fund. They advance three arguments in 14 support of their cause. 15 First, they argue that class counsel and lead plaintiff “lacked standing” to assert Section 16 11 claims, did not assert Section 11 claims, and yet settled those claims in violation of our 17 notice regarding class action settlement (Dkt. Nos. 50 at 3; 282 at 1, 2; 283 at 1). It is true that 18 our relevant prior order herein, dated November 21, 2019, advised the parties of the Court’s 19 expectations regarding class-action settlements: “The proposed release should be limited only 20 to the claims certified for class treatment” (ibid.). It also cautioned that in settlement, 21 defendants often seek to expand the class, either geographically (i.e., nationwide) or claim-wise (including claims not even in the 22 complaint) or person-wise (e.g., multiple new categories). Such expansions will be viewed with suspicion. If an expansion is to 23 occur it must come with an adequate plaintiff and one with standing to represent the add-on scope and with an amended complaint to 24 include the new claims, not to mention due diligence as to the expanded scope. The settlement dollars must be sufficient to cover 25 the old scope plus the new scope 26 (ibid. (emphasis added)). Our order granting plaintiffs’ unopposed motion certifying the class 27 indeed certified only Section 10(b) claims (Dkt. No. 127). Retirement System in fact lacks 1 our prior order, this order recognizes important mitigating considerations. Our court of 2 appeals’ binding precedent permits settlements such as these. See also Hesse, 598 F.3d at 591; 3 Class Plaintiffs v. City of Seattle, 955 F.2d 1268, 1288 (9th Cir. 1992). Intervenor counsel’s 4 objection that class counsel were not empowered to settle Section 10 claims for those also 5 having Section 11 claims was rejected at oral argument (see June 10, 2021 Hrg. Tr. at 37–38, 6 Dkt. No. 211). Intervenor counsel now again argue that Hesse requires “identical” facts in 7 order to settle all claims. They further argue that a stock offering and registration statement 8 represent “different” facts. Not so. The “factual” basis for each case lies in the identical 9 misrepresentations and omissions contained in the registration statement (see Dkt. No. 258 at 10 6). See In re Leapfrog Enterprises, Inc. Sec. Litig., 2005 WL 3801587, at *3 (N.D. Cal. Nov. 11 23, 2005) (Judge Ronald M. Whyte). 12 Second, they argue, the unopposed motion for class certification raised a red flag. True, 13 but a $129 million settlement can hardly be deemed collusive. 14 Third, intervenor counsel argue that they prosecuted their own suit in Santa Cruz County 15 Superior Court, that the settlement herein was announced two days after the state court’s 16 decision to deny defendants’ demurrer, and that a few days later the parties herein announced 17 their settlement, presumably indicating that the state suit helped to force the settlement. See 18 Nasseri v. Granite Constr. Inc., et al., Superior Ct. Santa Cruz County, 2021, No. 19CV03208; 19 Dkt. No. 124 at 5. Here is the timeline: According to intervenor counsel, a state-court hearing 20 was held on April 6, 2021, on the demurrer at which the court issued a tentative denial (June 21 15, 2021 Hrg. Tr. at 44:18–20, Dkt. No. 211). Further, on April 8, class counsel accepted 22 Judge Spero’s mediator’s offer (id. at 44–45:21–13). On April 9, 2021, defendants in the state 23 action officially lost their demurrer with an entry of a relevant order by the superior court. 24 Even if the state ruling was a factor, intervenor counsel claim far too much credit. The $129 25 million settlement flowed from the hard work of class counsel, the discovery they took, the 26 investigations they did, and their victories in court. The federal suit led to the restatement 27 (Dkt. No. 274-2 at 2). 1 In contrast, defendants won dismissal of Nasseri’s complaint in state court in mid-2020. 2 Then, Granite conducted its new audit and issued its restatement. Only then, in September 3 2020, did Nasseri’s amended complaint survive past the pleadings stage (ibid.). Intervenor 4 counsel were never appointed class counsel in the state case. It is not too much to say that 5 intervenor counsel’s suit plausibly would have failed but for the efforts of class counsel. To 6 repeat, in the Court’s judgment, the $129 million settlement is almost entirely the result of the 7 hard work of class counsel. 8 Only after class counsel obtained the $129 million settlement did intervenor counsel seek 9 a role herein. Now that class counsel has done all the hard work, intervenor counsel seek to 10 recover for all of their time litigating the state-court case, plus a multiplier, even though 11 nothing ever came to fruition in their suit. 12 Intervenor counsel do deserve credit for the incremental increase in recovery for those 13 class members with both Section 10(b) and Section 11 claims. That recovery went from 14 $15,549,968 in the original (preintervention) plan of allocation, to $37,555,065 under the 15 revised plan of allocation. Thus, overall, intervenor counsel prompted an increase in recovery 16 of $22,005,097 for the Section 10/11 claimants (Dkt. No. 224). But we must remember that 17 this increase was possible only after class counsel had already achieved the $129 million 18 overall settlement in the first place. The latter achievement stands as the monumental event 19 and the revision of the plan of allocation presupposed and was possible only because of that 20 achievement. Put differently, the recovery herein by those with both Section 10 and 11 claims 21 is fundamentally due to the work of class counsel, not intervenor counsel. 22 This order has considered the concrete benefit that intervenor counsel accomplished for 23 Section 10/11 claimants, and the comparative contribution. To repeat, the percentage award 24 for attorney’s fees will be 16.33% of the common fund after expenses are subtracted. 25 Intervenor counsel will recover $1,000,000 from the total attorney’s fees. The balance of 26 $19,921,494.16 will go to class counsel.3 27

C. LODESTAR CROSS-CHECK.

1 Our court of appeals has “encouraged courts to guard against an unreasonable result by 2 cross-checking their calculations against a second method.” Bluetooth, 654 F.3d at 944. The 3 lodestar figure multiples the number of hours reasonably expended on the litigation (as 4 supported by adequate documentation) by the reasonable hourly rate for the region and 5 experience of the lawyer. See id. at 941. 6 Intervenor counsel’s lodestar requires some modification. For their work from the 7 inception of the state case through the present, intervenor counsel claim a lodestar of 8 $1,807,481. However, the lodestar for intervenor counsel’s work on the settlement and plan of 9 allocation alone amounted to just $651,999 (Dkt. No. 274-1 Exh. B, General Project Nos. 6, 8, 10 10–14, 16, 19; Dkt. No. 280 at 7, n. 7, n. 8). 11 Billing rates for BFA and Robbins Geller, although on the high end in some respects, 12 qualify as reasonable. For the twelve partners who billed for this action, the hourly rate ranged 13 from Peter Borkon ($950) to Samuel H. Rudman and Darren J. Robbins ($1,325). The of 14 counsel and special counsel involved in this action, Ross Shikowitz and Sara Simnowitz, billed 15 $780. For the eleven associates, the rate ranged from John M. Kelley ($175) to Evan Kubota 16 ($690). Two summer associates billed: Stephen D. Johnson ($175) and Kyle Ezring ($330). 17 Nine staff attorneys participated herein, all billing either $390 or $395. Three investigations 18 and analyst staff participated, with their hourly rate ranging from Scott R. Roelen ($295) to 19 Andrew J. Rudolph ($750). An unknown number of paralegals assisted, with their billing rate 20 ranging from $275 to $350. Four support staff contributed, with rates ranging from Masiel 21 Feliz ($350) to Michael Russo ($445). An unknown number of document clerks participated at 22 the billing rate of $150. Finally, litigation support employee Bradley P. Lewis billed $150 23 (Alvarez Decl. Exh. A, Dkt. No. 274-1; Borkon Decl. Exh. B, Dkt. No. 277-2 at 2–6; Borkon 24 Decl. Exh. C, Dkt. No. 277-3 at 2). 25 26 27 1 This order accepts the claimed rates as generally tracking the going rate for those with 2 the same levels of skill and experience in our geographic region. See, e.g., In re Lidoderm 3 Antitrust Litig., 2018 WL 4620695, at *2 (N.D. Cal. Sept. 20, 2018) (Judge William H. 4 Orrick); Kudatsky v. Tyler, 2021 WL 5356724, at *5 (N.D. Cal. Nov. 17, 2021); Bisaccia v. 5 Revel Sys. Inc., 2019 WL 3220275, at *8 (N.D. Cal. July 17, 2019) (Judge Haywood Gilliam). 6 This order turns next to the reasonableness of billing entries. Counsel billed 12,512 7 hours (class counsel) and 2,127.3 hours (intervenor counsel), for a total of 14,639.3 hours in 8 this action (Borkon Decl. ¶ 106, Dkt. No. 277; Dkt. No. 277 at 17). 9 Hundreds of pages of time records detailed class counsel’s hours (Borkon Decl. Exh. E, 10 Dkt. No. 277-5). Intervenor counsel summarized the hours worked into twenty categories 11 visible on the public docket but filed the descriptions under seal (Alvarez Decl. Exh. B, Dkt. 12 No. 274-1; Dkt. No. 275-4). It was ill-advised to seek to seal these descriptions. By way of 13 explanation, intervenor counsel offered simply, “This breakdown provides a level of detail that 14 reveals specific research or litigation strategy which would be entitled to protection from 15 disclosure” (Alvarez Decl. ¶ 2, Dkt. No. 275-1 (cleaned up)). Attorney Alvarez’s reason falls 16 short of our Local Rules’ requirements and falls short of good cause. See Civ. L.R. 79-5. The 17 class deserves to know what justifications may exist for attorney’s fees that are siphoned from 18 their recovery. The motion to seal is DENIED. 19 Class counsel’s individual time entries appear sufficiently detailed, generally 20 comprehensible, and not overstated. This order also finds that class counsel’s timesheets do 21 not clearly indicate significant overstaffing, duplication, inefficiency, or too many meetings. 22 See Gutierrez v. Wells Fargo Bank, N.A., 2015 WL 2438274, at *6 (N.D. Cal. May 21, 2015). 23 Intervenor counsel objected to fees sought for class counsel’s three executed and thirteen 24 prepared depositions. Parties cancelled the thirteen once settlement got underway. Intervenor 25 counsel object to the line-item for those depositions, $1,960,279, as too high. The raw figure 26 appears acceptable for preparing and/or executing sixteen depositions (see Borkon Decl. ¶¶ 27 44–49, Dkt. No. 277). 1 Intervenor counsel further object that class counsel seek $1.2 million for an unopposed 2 class certification motion (Dkt. No. 282 at 3). In reality, that total represents roughly $391,969 3 times 2.97, class counsel’s original proposed multiplier (Dkt. Nos. 282 at 2, 4; 277-4 at 3; 284 4 at 5). Moreover, Rule 23 assigns to the district court a supervisory role. Even without an 5 opposition, class counsel had to put in the work so that the district court could certify that the 6 putative class met Rule 23’s requirements. Furthermore, class counsel hired a third party to 7 review its lodestar, which resulted in some deletions or modifications. Both points strengthen 8 the reliability of class counsel’s billing. Class counsel’s net lodestar was $7,079,129 (Borkon 9 Decl. ¶ 106, Dkt. No. 277). 10 Granting all but $1,000,000 of the fee award to class counsel amounts to a total recovery 11 of $19,921,494.16, for a multiplier of 2.81. 12 Intervenor counsel’s billing raises red flags. Many of Attorney Xavier J. Alvarez’s 13 entries say simply, “review and analyze” documents, such as SEC filings, briefings, or 14 caselaw. More concerning, this order finds the risk of inflated billing unusually high, given 15 that twelve people each billed fewer than sixteen hours. See Gutierrez, 2015 WL 2438274, at 16 *6. (Alvarez Decl. Exh. A, Dkt. No. 274-1). Class counsel omitted any individuals’ hours 17 from calculation if that person worked less than twenty-five hours on the case (Feb. 24, 2022 18 Hrg. Tr. at 15:13–15, Dkt. No. 297). This makes sense. Even sixteen hours on a case would 19 amount to roughly one day’s worth of work at an average firm. This time would not permit an 20 attorney to contribute meaningfully to the case. This order reduces intervenor counsel’s total 21 lodestar by the amount associated with the five attorneys whose hours totaled less than twenty- 22 five, both for parity and to reduce inflationary billing. All for the purpose of a lodestar cross- 23 check, intervenor counsel’s adjusted lodestar hours clock in at 2,083.5, for a total adjusted 24 lodestar number of $1,760,826.50.4 25

26 4 This order subtracts hours for attorneys Joseph D. Daley (12.3 hours at $970/hour), James I. Jaconette (15.4 hours at $995/hour), Darren J. Robbins (2.5 hours at $1,325/hour), Samuel H. 27 Rudman (11.2 hours at $1,325/hour), and Matthew J. Balotta (2.4 hours at $520/hour), for a total 1 Class counsel also request that this order discount some of intervenor counsel’s lodestar 2 since it relates to pre-settlement work, which “was of questionable value” based on the fact that 3 the state action grounded much of its successful complaint on the federal amended complaint 4 herein (Dkt. No. 284 at 2, citing Dkt. No. 280 at 6). The other half of intervenor counsel’s 5 lodestar, according to class counsel, accrued post-settlement, when the risk of nonrecovery was 6 “practically zero.” Furthermore, as stated, intervenor counsel’s lodestar number for work on 7 the settlement and the plan of allocation was $651,999 (Dkt. Nos. 280 at 9; 274-1, Exh. B, 8 General Project Nos. 6, 8, 10–14, 16, 19). 9 The critiques have merit. Intervenor counsel spent 744.4 hours on the plan of allocation 10 and issues related to settlement, while class counsel spent 872.8 hours on the settlement as a 11 whole, which also included the notice and claims process (Dkt. No. 280 at 7). This point 12 strengthens concerns about inflated billing. 13 Our award to intervenor counsel, $1,000,000, yields a multiplier of 0.57 on the adjusted 14 lodestar number (see, supra, n. 4). This is reasonable since intervenor counsel entered late in 15 the game and contributed meaningfully but minimally. For the work on the settlement and 16 plan of allocation only, this order has not surveyed the billing belonging to intervenor 17 counsel’s five lawyers who worked less than twenty-five hours to tell what projects each 18 worked on. Suffice it to say that it is possible that none of the $46,654.50 spent on those five 19 lawyers occurred during the projects related to the plan of allocation and settlement herein, or 20 all of it did (see, supra, n. 4). Adjusting the lodestar for that work from $651,999 by the entire 21 $46,654.50 and, next, comparing that figure to the assumption that none of these lawyers 22 worked on the plan of allocation or settlement herein, yields a range of adjusted lodestars of: 23 $605,345 – $651,999. Thus, our award to intervenor counsel provides a multiplier range for 24 the settlement and plan of allocation work of 1.53 – 1.65. This is reasonable. See Indep. 25 Living Ctr. of S. California v. Kent, 2020 WL 418947, at *2 (C.D. Cal. Jan. 24, 2020) (Judge 26 Christina A. Snyder), rev’d on other grounds Indep. Living Ctr. of S. California, Inc. v. 27 1 Figueroa, 856 F. App’x 97 (9th Cir. 2021) (awarding 1.5 multiplier for intervenor counsel 2 with, inter alia, ten years in the case).5 3 All multipliers discussed above track the acceptable range commonly applied in our 4 circuit. See, e.g., In re Hyundai & Kia Fuel Economy Litig., 926 F.3d 539, 571–72 (9th Cir. 5 2019); Vizcaino, 290 F.3d at 1051 n. 6 (describing appendix to opinion, finding a range of 6 multipliers in common fund cases of 0.6–19.6). 7 To summarize, class counsel’s multiplier on a recovery of $19,921,494.16 amounts to 8 2.81. Intervenor counsel’s multiplier on a recovery of $1,000,000 amounts to 0.57 on their 9 total lodestar and range of 1.53 – 1.65 on the lodestar for their work on the settlement and the 10 plan of allocation. 11 D. INCENTIVE AWARD. 12 Intervenor counsel request $12,875 for Nasseri’s service herein and in the state-court 13 action, where he acted as named plaintiff (Dkt. 274 at 19–20). Class counsel oppose. The 14 award of incentive fees can “entice” a representative to support meager settlements. In re 15 Diamond Foods, 2014 WL 106826, at *4; see also Staton v. Boeing, 327 F.3d 938, 977 (9th 16 Cir. 2003). In the state-court action, Nasseri did not achieve class certification, so he only ever 17 represented himself. Furthermore, he will personally benefit from the intervention (Nasseri 18 Decl., Dkt. No. 274-3). His willingness to serve is appreciated, but the request is DENIED. 19 Objections to his declaration are DENIED AS MOOT. 20 E. EXPENSES. 21 Class counsel seek $763,958 in total, an amount encompassing expenses for experts, a 22 discovery document database, filing fees, copying, postage, and other unsurprising line items 23 (Borkon Decl. ¶¶ 130–39, Dkt. No. 277; id. Exh. I, Dkt. No. 277-9). Understandably, the 24 25

26 5 The range of multipliers was calculated as follows. Multiplier on intervenor counsel’s unadjusted lodestar on work for settlement and plan of allocation only: $1,000,000 / $651,999 = 27 1.53. Multiplier on intervenor counsel’s lodestar on work for settlement and plan of allocation 1 majority ($721,772) went to a combination of expert fees and litigation-support vendor fees.6 2 Counsel seek less than $3,000 for out-of-town transportation and less than $500 for 3 accommodations, all for travel that occurred prior to the pandemic (ibid.). We received no 4 objections to this request, and Retirement System supports. The expenses are reasonable. 5 Intervenor counsel seek $119,115.13. Their expert costs exceeded all others, and Nasseri 6 supports (Alvarez Decl. ¶ 8(c), Exh. D, Dkt. No. 274-1). 7 All expenses appear reasonable. 8 CONCLUSION 9 To the extent stated, final approval is GRANTED. Intervenor counsel’s motion to seal 10 (Dkt. No. 275) is DENIED. All of plaintiff’s claims are hereby dismissed with prejudice as 11 provided in the settlement agreement. The request that this order find the notice of settlement 12 adequate and satisfactory under Rule 23, the PLSRA, and due process, is GRANTED. The 13 request for final approval of the plan of allocation for the settlement is GRANTED. The request 14 that the Court retain exclusive jurisdiction over the settlement and the settlement agreement as 15 described therein and over this order and final judgment is GRANTED FOR A PERIOD OF SIX 16 MONTHS FROM ENTRY OF JUDGMENT. The further request for authorization for class counsel 17 and Epiq to begin administration and distribution of the net proceeds of the settlement in 18 accordance with the plan of allocation is GRANTED. 19 Class counsel and intervenor counsel are AWARDED attorney’s fees in the amount of 20 $20,921,494.16, plus any interest on that amount that may accrue before distribution. Half of 21 the award shall be distributed as of the effective date as defined in the settlement agreement. 22 The remaining half shall be paid when both sides certify that all funds have been properly 23 distributed and the file can be completely closed. At all times, counsel will split the fee award 24 as described herein. Class counsel and intervenor counsel are AWARDED expenses in the 25 26 27 1 amounts of $763,958 to BFA and $119,115.13 to Robbins Geller. 2 IT IS SO ORDERED. 3 Dated: March 17, 2022. 4 Pee 5 [ A =

WILLIAM ALSUP

6 UNITED STATES DISTRICT JUDGE 7 8 9 10 11 a 12

13 14

15 16

= 17

Z 18

19 20 21 22 23 24 25 26 27 28