Golden v. BofI Holding, Inc. et al

Filing 247

ORDER Granting Plaintiff's 205 Motion for Class Certification, Appointment of Class Representative, and Approval of Class Counsel. Signed by Judge Gonzalo P. Curiel on 8/23/21. (dlg)

Case 3:15-cv-02324-GPC-KSC Document 247 Filed 08/24/21 PageID.5867 Page 1 of 21

1
2
3
4
5
6
7
8

UNITED STATES DISTRICT COURT

9

SOUTHERN DISTRICT OF CALIFORNIA

10
11
12

IN RE:
BofI HOLDING, INC. SECURITIES
LITIGATION.

Case No.: 3:15-cv-02324-GPC-KSC
ORDER GRANTING PLAINTIFF’S
MOTION FOR CLASS
CERTIFICATION, APPOINTMENT
OF CLASS REPRESENTATIVE,
AND APPROVAL OF CLASS
COUNSEL

13
14
15
16
17

[ECF No. 205]

18
On May 28, 2021, Lead Plaintiff Houston Municipal Employees Pension System

19
20
21
22
23
24
25
26
27
28

filed a Motion for Class Certification, Appointment of Class Representative, and
Approval of Class Counsel. ECF No. 205. Defendants BofI Holding, Inc., Gregory
Garrabrants, Andrew J. Micheletti, Paul J. Grinberg, Nicholas A. Mosich and James S.
Argalas oppose. ECF No. 211. The motion is fully briefed. ECF Nos. 205, 211, 226.
On August 20, 2021, the Court held a hearing on this matter. ECF No. 245. Upon
consideration of the briefing and arguments of the parties and for the reasons set forth
below, the Court GRANTS the motion.
\\\
\\\
1
3:15-cv-02324-GPC-KSC

Case 3:15-cv-02324-GPC-KSC Document 247 Filed 08/24/21 PageID.5868 Page 2 of 21

1

I.

2

Background
This case is a consolidated putative securities fraud class action brought by

3

purchasers of BofI’s1 stock for violations of Sections 10(b) and 20(a) of the Securities

4

Exchange Act of 1934. On February 1, 2016, the Court appointed Houston Municipal

5

Employees Pension System as Lead Plaintiff (“Plaintiff” or “HMEPS”). ECF No. 23.

6

The operative pleading in this case is the Third Amended Complaint (the “TAC”).

7

ECF No. 136. On March 21, 2018, the Court granted Defendants’ motion to dismiss the

8

TAC with prejudice. ECF No. 156. The Court concluded that the TAC failed to identify

9

a corrective disclosure of the alleged misrepresentations with the particularity required by

10

Federal Rule of Civil Procedure (“Rule”) 9(b). Id. at 9.2 Specifically, the Court

11

determined that the two alleged corrective disclosures—the complaint in Erhart v. BofI

12

Holding, Inc., No. 3:15-cv-02287-BAS-NLS (S.D. Cal.), ECF No. 1 (the “Erhart

13

Complaint”)3 filed against BofI by Charles Matthew Erhart, a former BofI internal

14

auditor, and several articles by Seeking Alpha—could not establish loss causation. With

15

respect to the Erhart Complaint, the Court found that the complaint was at most a

16

“partial” corrective disclosure of Defendants’ misrepresentations about BofI’s internal

17

controls because the allegations, standing alone, did not confirm the fraud. Id. at 14.

18

Plaintiff appealed, and the Ninth Circuit reversed. In re BofI Holding, Inc. Sec.

19

Litig., 977 F.3d 781 (9th Cir. 2020). The Ninth Circuit agreed that the Seeking Alpha

20

articles could not constitute corrective disclosures, though for a slightly different reason,

21

but determined that Plaintiff adequately alleged the Erhart Complaint was a corrective

22

disclosure and that the loss causation element was satisfied. Id. at 786, 794, 797. The

23
24
25
26
27
28

“BofI is the holding company for BofI Federal Bank, a federally chartered savings association that
purportedly operates from its single location in San Diego.” TAC, ECF No. 136 at ¶ 28. The entities
now operate under a different corporate name, but the Court will continue to use “BofI” to refer to both
the holding company and its subsidiary BofI Federal Bank.
2
References to page numbers follow the CM/ECF pagination.
3
The Court GRANTS Defendants’ request for judicial notice of the Erhart Complaint. Fed. R. Evid.
201(b); see also In re BofI Holding, Inc. Sec. Litig., 977 F.3d 781, 791 n.2 (9th Cir. 2020).
1

2
3:15-cv-02324-GPC-KSC

Case 3:15-cv-02324-GPC-KSC Document 247 Filed 08/24/21 PageID.5869 Page 3 of 21

1

court noted that the shareholders did not need to establish that the allegations in the

2

Erhart Complaint are true; rather, “the relevant question for loss causation purposes is

3

whether the market reasonably perceived Erhart’s allegations as true and acted upon them

4

according.” Id. at 791–92. The Ninth Circuit’s order ultimately left in the case two

5

categories of misstatements—concerning (1) the bank’s underwriting standards and (2)

6

its system of internal controls and compliance infrastructure—and the Erhart Complaint

7

as the only potential corrective disclosure. Id. at 786–87, 798.

8

Plaintiff has now filed a motion for class certification under Federal Rule of Civil

9

Procedure 23(b)(3), appointment of HMEPS as class representative, and appointment of

10

Lieff Cabraser Heimann & Bernstein, LLC (“Lieff Cabraser”) as class counsel. ECF No.

11

205. The proposed class definition is “all persons and entities that, during the Class

12

Period, purchased or otherwise acquired shares of the publicly traded common stock of

13

BofI, as well as purchasers of BofI call options and sellers of BofI put options, and were

14

damaged thereby.” ECF No. 205-1 at 6. Plaintiff defines the Class Period as running

15

from September 4, 2013, through October 14, 2015, inclusive. Id. at 6, n.2.

16
17

II.

Legal Standard
“The class action is an exception to the usual rule that litigation is conducted by

18

and on behalf of individual named parties only. In order to justify a departure from that

19

rule, a class representative must be a part of the class and possess the same interest and

20

suffer the same injury as the class members.” Wal-Mart Stores, Inc. v. Dukes, 564 U.S.

21

338, 348 (2011) (internal quotation marks and citations omitted). Accordingly, “a party

22

seeking to maintain a class action ‘must affirmatively demonstrate his compliance’ with

23

[Federal Rule of Civil Procedure] 23.” Comcast Corp. v. Behrend, 569 U.S. 27, 33

24

(2013) (quoting Dukes, 564 U.S. at 350–51).

25

Rule 23 contains two sets of requirements. First, “Rule 23(a) ensures that the

26

named plaintiffs are appropriate representatives of the class whose claims they wish to

27

litigate. The Rule’s four requirements—numerosity, commonality, typicality, and

28

adequate representation—effectively limit the class claims to those fairly encompassed
3
3:15-cv-02324-GPC-KSC

Case 3:15-cv-02324-GPC-KSC Document 247 Filed 08/24/21 PageID.5870 Page 4 of 21

1

by the named plaintiff’s claims.” Dukes, 564 U.S. at 349 (internal quotation marks and

2

citations omitted). Second, “[w]here a putative class satisfies all four requirements of

3

23(a), it still must meet at least one of the three additional requirements outlined in

4

23(b).” United Steel, Paper & Forestry, Rubber, Mfg. Energy, Allied Indus. & Serv.

5

Workers Int’l Union AFL-CIO, CLC v. ConocoPhillips Co., 593 F.3d 802, 806 (9th Cir.

6

2010).

7

To proceed as a class action, plaintiffs “must actually prove—not simply plead—

8

that their proposed class satisfies each requirement of Rule 23.” Halliburton Co. v. Erica

9

P. John Fund, Inc., 573 U.S. 258, 275 (2014) (“Halliburton II”). But on a motion for

10

class certification, the Court is required to “examine the merits of the underlying claim . .

11

. only inasmuch as it must determine whether common questions exist; not to determine

12

whether class members could actually prevail on the merits of their claims.” Ellis v.

13

Costco Wholesale Corp., 657 F.3d 970, 981 n.8 (9th Cir. 2011) (citations omitted).

14

III.

Discussion

15

Defendants do not “affirmatively dispute” that Plaintiff meets the Rule 23(a)

16

requirements, and instead focus on the predominance requirement imposed by Rule

17

23(b)(3). ECF No. 211 at 14. However, the Court will independently evaluate whether

18

Plaintiff meets the prerequisites to certification. See Davis v. Hutchins, 321 F.3d 641,

19

649 (7th Cir. 2003) (“Rule 23(c) imposes an independent duty on the district court to

20

determine by order that the requirements of Rule 23(a) are met.”).

21
22

a. Rule 23(a)
Federal Rule of Civil Procedure (“Rule”) 23 establishes four prerequisites for class

23

certification: (1) numerosity; (2) commonality; (3) typicality; and (4) adequacy of

24

representation. Fed. R. Civ. P. 23(a). The Court considers each in turn.

25
26

1. Numerosity
The numerosity requirement under Rule 23(a)(1) is met if “the class is so

27

numerous that joinder of all members is impracticable.” Fed. R. Civ. P. 23(a)(1). “As a

28

general matter, courts have found that numerosity is satisfied when class size exceeds 40
4
3:15-cv-02324-GPC-KSC

Case 3:15-cv-02324-GPC-KSC Document 247 Filed 08/24/21 PageID.5871 Page 5 of 21

1

members, but not satisfied when membership dips below 21.” Slaven v. BP Am., Inc.,

2

190 F.R.D. 649, 654 (C.D. Cal. 2000). Because BofI had millions of shares trading on

3

NASDAQ during the Class Period, see ECF No. 205-3 (Report of Frank C. Torchio or

4

“Torchio Report”) ¶ 14, the Court can infer that the number of shareholders and other

5

potential class members damaged by Defendants’ actions would be far too numerous to

6

join. See Howell v. JBI, Inc., 298 F.R.D. 649, 654 (D. Nev. 2014) (“[I]n securities cases,

7

when millions of shares are traded during the proposed class period, a court may infer

8

that the numerosity requirement is satisfied.”); Blackie v. Barrack, 524 F.2d 891, 901 (9th

9

Cir. 1975) (noting that “understandably,” defendants did not challenge compliance with

10

the numerosity requirement where “it appears that the class period . . . will encompass the

11

purchasers involved in about 120,000 transactions involving some 21,000,000 shares”).

12

Accordingly, the numerosity requirement is met.

13
14

2. Commonality
Rule 23(a)(2) requires the existence of “questions of law or fact common to the

15

class.” Fed. R. Civ. P. 23(a)(2). Commonality is established if plaintiffs and class

16

members’ claims “depend upon a common contention . . . capable of class-wide

17

resolution—which means that determination of its truth or falsity will resolve an issue

18

that is central to the validity of each one of the claims in one stroke.” Dukes, 564 U.S. at

19

350. “[A] class meets Rule 23(a)(2)’s commonality requirement when the common

20

questions it has raised are apt to drive the resolution of the litigation, no matter their

21

number.” Jimenez v. Allstate Ins. Co., 765 F.3d 1161, 1165 (9th Cir. 2014) (internal

22

quotation marks and citation omitted). While Defendants contest the predominance of

23

common issues in this case under Rule 23(b)(3), it is clear that the case exceeds Rule

24

23(a)(2)’s minimum requirement that there be “[e]ven a single [common] question.”

25

Dukes, 564 U.S. at 359 (alteration in original). Even excluding damages and loss

26

causation issues, which are the main subject of the parties’ Rule 23(b)(3) predominance

27

dispute, several of the elements of the securities fraud claim here are indisputably subject

28

to common proof and identical legal analysis. Common questions include whether
5
3:15-cv-02324-GPC-KSC

Case 3:15-cv-02324-GPC-KSC Document 247 Filed 08/24/21 PageID.5872 Page 6 of 21

1

Defendants’ actions violated federal securities laws; whether the statements identified in

2

the TAC constitute material misrepresentations; and whether Defendants acted with the

3

requisite scienter. See Amgen Inc. v. Connecticut Ret. Plans & Tr. Funds, 568 U.S. 455,

4

460–61 (2013) (citation omitted) (noting elements of a private securities-fraud action); cf.

5

In re Juniper Networks, Inc. Sec. Litig., 264 F.R.D. 584, 588–89 (N.D. Cal. 2009) (noting

6

that the presence of the same common issues satisfied Rule 23(a)(2)). The Court

7

therefore finds that this case meets the commonality requirement imposed by Rule

8

23(a)(2).

9

3. Typicality

10

Rule 23(a)(3)’s typicality requirement will be satisfied when “the claims or

11

defenses of the representative parties are typical of the claims or defenses of the class.”

12

Fed. R. Civ. P. 23(a)(3). The named plaintiff must be a member of the class they seek to

13

represent and must “possess the same interest and suffer the same injury” as putative

14

class members. Gen. Tel. Co. of Sw. v. Falcon, 457 U.S. 147, 156 (1982) (internal

15

quotations omitted). Here, HMEPS’s claim is essentially the same as that of the

16

proposed class members: HMEPS purchased stock, call, or put options at what it alleges

17

to be an artificially inflated or deflated price as a result of material misrepresentations

18

made by Defendants, and then suffered damages as a result of the corrective disclosure.

19

Defendants do not assert any defenses unique to HMEPS. See Sudunagunta v.

20

NantKwest, Inc., No. CV161947MWFJEMX, 2018 WL 3917865, at *4 (C.D. Cal. Aug.

21

13, 2018) (quoting William B. Rubenstein, Newberg on Class Actions, § 22:72 (5th ed.

22

2018)) (noting that securities claims tend to avoid typicality problems such as defenses

23

unique to the named plaintiff). The Court thus concludes that HMEPS’s claims are

24

typical of that of the class.

25
26

4. Adequacy
Under Rule 23(a)(4), representative parties must be able to “fairly and adequately

27

protect the interests of the class.” Fed. R. Civ. P. 23(a)(4). In analyzing whether Rule

28

23(a)(4) has been met, the Court must ask two questions: “(1) do the named plaintiffs and
6
3:15-cv-02324-GPC-KSC

Case 3:15-cv-02324-GPC-KSC Document 247 Filed 08/24/21 PageID.5873 Page 7 of 21

1

their counsel have any conflicts of interest with other class members and (2) will the

2

named plaintiffs and their counsel prosecute the action vigorously on behalf of the

3

class?” Evon v. Law Offices of Sidney Mickell, 688 F.3d 1015, 1031 (9th Cir. 2012)

4

(citation omitted). The Court must also consider the adequacy of counsel before

5

appointing class counsel under Rule 23(g). Fed. R. Civ. P. 23(g); Scholl v. Mnuchin, 489

6

F. Supp. 3d 1008, 1045 (N.D. Cal. 2020).

7

HMEPS has demonstrated the ability to litigate this securities-fraud action

8

vigorously on behalf of class members and does not appear to have any interests adverse

9

to those of the class. Indeed, as a large institutional investor that purchased a large

10

financial interest in BofI, see ECF No. 205-4 (Declaration of Sherry Mose or “Mose

11

Decl.”) ¶ 3, HMEPS is the type of plaintiff typically favored under the PSLRA to be the

12

lead plaintiff in securities litigation. See 15 U.S.C.A. § 78u-4(a)(3)(B)(iii)(I). HMEPS

13

likewise satisfies the separate adequacy requirement under Rule 23(a)(4) for appointment

14

as class representative because it has actively participated in the litigation of this case, see

15

Mose Decl. ¶¶ 4 –7, and there is no indication that its interests are in conflict with other

16

Class Period investors. The Court therefore determines that HMEPS is able to “fairly and

17

adequately protect the interests of the class” and can be appointed as class representative.

18

Fed. R. Civ. P. 23(a)(4).

19

The Court also finds that Lieff Cabraser satisfies both the adequacy requirement

20

and the prerequisites to appointment as class counsel under Rule 23(g). Lieff Cabraser

21

has litigated this case over the past five years, both before this Court and the Ninth

22

Circuit. See Fed. R. Civ. P. 23(g)(1)(A)(i), (iv); Mose Decl. ¶ 8. Counsel has significant

23

experience with securities fraud class actions and demonstrates a thorough understanding

24

of the applicable law. See Fed. R. Civ. P. 23(g)(1)(A)(ii), (iii); ECF No. 205-5 (Lieff

25

Cabraser Firm Resume) at 4–5. There is also no indication that the firm’s interests are

26

adverse to those of the class. Thus, the Court finds that Lieff Cabraser will fairly and

27

adequately represent the interests of the class and can be appointed as class counsel. See

28

Fed. R. Civ. P. 23(g)(2).
7
3:15-cv-02324-GPC-KSC

Case 3:15-cv-02324-GPC-KSC Document 247 Filed 08/24/21 PageID.5874 Page 8 of 21

1
2

b. Rule 23(b)(3)
Plaintiff seeks to maintain this class action under Rule 23(b)(3), which requires

3

that “questions of law or fact common to class members predominate over any questions

4

affecting only individual members, and that a class action is superior to other available

5

methods for fairly and efficiently adjudicating the controversy.” Fed. R. Civ. P. 23(b)(3).

6

Defendants’ opposition is primarily focused on the predominance requirement.

7
8

1. Predominance
As the Supreme Court recognized, “[i]n securities class action cases, the crucial

9

requirement for class certification will usually be the predominance requirement of Rule

10

23(b)(3).” Halliburton II, 573 U.S. at 276. The Court therefore must consider whether

11

questions capable of resolution with “generalized, class-wide proof” predominate over

12

individualized ones. See Tyson Foods, Inc. v. Bouaphakeo, 577 U.S. 442, 453 (2016).

13

“Predominance is not, however, a matter of nose-counting. Rather, more important

14

questions apt to drive the resolution of the litigation are given more weight in the

15

predominance analysis over individualized questions which are of considerably less

16

significance to the claims of the class.” In re Hyundai & Kia Fuel Econ. Litig., 926 F.3d

17

539, 557 (9th Cir. 2019) (quoting Ruiz Torres v. Mercer Canyons Inc., 835 F.3d 1125,

18

1134 (9th Cir. 2016)).

19

The Court’s predominance inquiry begins with the elements of a Section 10(b)

20

securities fraud claim, which are: “(1) a material misrepresentation or omission by the

21

defendant; (2) scienter; (3) a connection between the misrepresentation or omission and

22

the purchase or sale of a security; (4) reliance upon the misrepresentation or omission; (5)

23

economic loss; and (6) loss causation.” Erica P. John Fund, Inc. v. Halliburton Co., 563

24

U.S. 804, 809–10 (2011) (“Halliburton I”) (citations omitted). Defendants do not dispute

25

that several of the elements of Plaintiff’s securities fraud claim would be common to the

26

class. Indeed, the materiality of the misrepresentation and the defendant’s scienter are

27

issues that would require the same proof for any class member. See Amgen, 568 U.S. at

28

467; Petrie v. Elec. Game Card, Inc., 308 F.R.D. 336, 344 (C.D. Cal. 2015).
8
3:15-cv-02324-GPC-KSC

Case 3:15-cv-02324-GPC-KSC Document 247 Filed 08/24/21 PageID.5875 Page 9 of 21

1
2

i. Reliance
“Whether common questions of law or fact predominate in a securities fraud action

3

often turns on the element of reliance.” Halliburton I, 563 U.S. at 810. Although the

4

traditional way of demonstrating reliance is by showing that an individual purchaser of a

5

security, for example, was aware of the statement or omission and entered into the

6

transaction because of the statement or omission, in most securities cases requiring proof

7

of such direct reliance “would place an unnecessarily unrealistic evidentiary burden on

8

the Rule 10b–5 plaintiff who has traded on an impersonal market.” Basic Inc. v.

9

Levinson, 485 U.S. 224, 245 (1988). Accordingly, the Supreme Court in Basic held that

10

plaintiffs can invoke a rebuttable presumption of reliance based on the “‘fraud-on-the-

11

market’ theory, which holds that ‘the market price of shares traded on well-developed

12

markets reflects all publicly available information, and, hence, any material

13

misrepresentations.’” Halliburton II, 573 U.S. at 268 (quoting Basic, 485 U.S. at 246).

14

To obtain the presumption of reliance, a plaintiff must show “(1) that the alleged

15

misrepresentations were publicly known, (2) that they were material, (3) that the stock

16

traded in an efficient market, and (4) that the plaintiff traded the stock between the time

17

the misrepresentations were made and when the truth was revealed.” Id. (quoting Basic,

18

485 U.S. at 248, n.27). The materiality of the statements need not be proven at the class

19

certification stage. Amgen, 568 U.S. at 460.

20

Here, the misrepresentations at issue were included in SEC filings and proxy

21

statements, and were thus publicly known. See In re Banc of California Sec. Litig., 326

22

F.R.D. 640, 648 (C.D. Cal. 2018). Both HMEPS and members of the putative class made

23

the relevant transactions after at least some misrepresentations were made and before the

24

corrective disclosures. Defendants agreed not to challenge the efficiency of the market

25

for the purposes of this motion but argue that Plaintiff must still put forth evidence to

26

establish this element. ECF No. 205-2 (Declaration of Richard M. Heimann) ¶ 2; ECF

27

No. 211 at 14–15 n.4. The Court finds that Plaintiff has adequately done so. BofI stock

28

traded on NASDAQ, a well-developed market, and Mr. Torchio’s reply report explains
9
3:15-cv-02324-GPC-KSC

Case 3:15-cv-02324-GPC-KSC Document 247 Filed 08/24/21 PageID.5876 Page 10 of 21

1

how BofI stock satisfied the factors from Cammer v. Bloom, which many courts have

2

found helpful in determining market efficiency. ECF No. 226-3 (Reply Report of Frank

3

C. Torchi or “Torchio Reply Report”) ¶ 4, Appendix A; Binder v. Gillespie, 184 F.3d

4

1059, 1065 (9th Cir. 1999) (citing Cammer v. Bloom, 711 F. Supp. 1264, 1286–87

5

(D.N.J. 1989)). Accordingly, Plaintiff can invoke the presumption of reliance based on

6

the fraud-on-the-market theory, rendering reliance a common issue.

7
8

ii. Loss Causation
Loss causation “requires a plaintiff to show that a misrepresentation that affected

9

the integrity of the market price also caused a subsequent economic loss.” Halliburton I,

10

563 U.S. at 812 (emphasis in original). The Supreme Court has held that plaintiffs need

11

not establish loss causation at the class certification stage. Halliburton I, 563 U.S. at 813.

12

Although Halliburton I was confronted with the question of whether loss causation must

13

be established to invoke the rebuttable presumption of reliance under Basic, see id., it

14

also makes sense that loss causation typically will not cause a hurdle to class certification

15

under Rule 23(b)(3) because the issue is capable of resolution on a class-wide basis.

16

When loss causation is sought to be proved via the effect of a corrective disclosure, all

17

plaintiffs will need to show that the revelation of the misrepresentations caused the stock

18

price to drop from its artificially inflated level. In re BofI, 977 F.3d at 789–90 (citing

19

Dura Pharmaceuticals, Inc. v. Broudo, 544 U.S. 336, 347 (2005)). Whether the

20

corrective disclosure led to a reduction in stock price does not require any individualized

21

determinations that would vary across the class, particularly in a case like this one where

22

there is a single corrective disclosure. The Court therefore finds that loss causation is an

23

issue common to all putative class members’ claims.

24

iii. Damages

25

To meet the predominance requirement under Rule 23(b)(3), “plaintiffs must be

26

able to show that their damages stemmed from the defendant’s actions that created the

27

legal liability” under the proposed damages model. Leyva v. Medline Indus. Inc., 716

28

F.3d 510, 514 (9th Cir. 2013) (citing Comcast, 569 U.S. at 38). The Court must conduct
10
3:15-cv-02324-GPC-KSC

Case 3:15-cv-02324-GPC-KSC Document 247 Filed 08/24/21 PageID.5877 Page 11 of 21

1

a “rigorous analysis” to determine whether the model damages model is consistent with

2

the plaintiffs’ case for liability, although “[c]alculations need not be exact.” Comcast,

3

569 U.S. at 35 (citations omitted). “[U]ncertainty regarding class members’ damages

4

does not prevent certification of a class as long as a valid method has been proposed for

5

calculating those damages.” Nguyen v. Nissan N. Am., Inc., 932 F.3d 811, 817 (9th Cir.

6

2019) (quoting Lambert v. Nutraceutical Corp., 870 F.3d 1170, 1182 (9th Cir. 2017),

7

rev’d on other grounds, 139 S. Ct. 710 (2019)).

8
9

Defendants contend that Plaintiff has not met the predominance requirement
because it has not demonstrated a reliable method of class-wide calculation of damages

10

that aligns with the theory of liability as required by Comcast Corp. v. Behrend, 569 U.S.

11

27 (2013). Plaintiff argues that its proposed damages model is standard in securities class

12

actions and has been found by numerous courts to satisfy the requirements of Comcast,

13

and that Defendants’ argument instead improperly seeks to force Plaintiff to prove loss

14

causation at the class certification stage.

15

In Comcast, an antitrust case, the Supreme Court found that deficiencies in the

16

plaintiff’s proposed damages model precluded certification of a class under Rule

17

23(b)(3). The plaintiff, seeking to represent cable subscribers across Philadelphia, had

18

asserted four theories of antitrust violations against cable provider Comcast, but the

19

district court found that only one of the four theories—that Comcast’s actions deterred

20

competition from “overbuilders”—could proceed on a class-wide basis. Id. at 31.

21

However, the district court found that damages could be calculated on class-wide basis as

22

to the remaining theory of liability, even though the proposed model of calculating

23

damages accounted for the harm caused by all four alleged antitrust violations and did not

24

isolate damages caused by overbuilding deterrence. Id. at 32, 36–37. The Third Circuit

25

affirmed, and refused to consider arguments related to the damages model because they

26

constituted an improper attack on the merits. Id. at 32.

27
28

After conducting the “rigorous analysis” of the damages model that the Third
Circuit had failed to undertake, the Supreme Court determined that its deficiencies were
11
3:15-cv-02324-GPC-KSC

Case 3:15-cv-02324-GPC-KSC Document 247 Filed 08/24/21 PageID.5878 Page 12 of 21

1

fatal to class certification under Rule 23(b)(3). Id. at 33–34. The Court noted that

2

despite the Third Circuit’s reference to the requirement that the plaintiff must “assure [the

3

court] that if they can prove antitrust impact, the resulting damages are capable of

4

measurement and will not require labyrinthine individual calculations,” the proposed

5

damages model indeed left open the possibility that measuring the harm caused by the

6

remaining theory of liability would not be possible on a class-wide basis. Id. at 37

7

(quoting Behrend v. Comcast Corp., 655 F.3d 182, 206 (3d Cir. 2011)). As the Court

8

explained, cable subscribers in different counties may have been overcharged as a result

9

of different forms of anticompetitive conduct, making it impossible to calculate damages

10

attributable solely to overbuilding deterrence with the existing proposed model. Id. at

11

37–38. Because the damages model was unable to account for the “nearly endless”

12

“permutations involving four theories of liability and 2 million subscribers located in 16

13

counties,” the Court held that “Rule 23(b)(3) cannot authorize treating subscribers within

14

the Philadelphia cluster as members of a single class.” Id. at 38.

15

Here, Plaintiff provides the Report of Frank C. Torchio, ECF No. 205-3, to

16

describe the method Plaintiff plans to use to calculate damages on a class-wide basis.

17

Mr. Torchio proposes to use a model known as an event study to measure class members’

18

economic losses stemming from their acquisition of the security or related interest at an

19

artificially inflated price that was later reduced once the misrepresentations were revealed

20

to the market. Torchio Report ¶ 17. In general terms, an event study uses a regression

21

analysis to determine the effect of an event (such as a corrective disclosure) on a

22

dependent variable (such as the share price). See Torchio Report ¶¶ 23–41; In re Imperial

23

Credit Indus., Inc. Sec. Litig., 252 F.Supp.2d 1005, 1014 (C.D. Cal. 2003). Defendants

24

contend that the proposed damages model does not align with Plaintiff’s theory of

25

liability, as required under Comcast, because it fails to account for the possible falsity of

26

allegations in the Erhart Complaint, the impact of non-disclosed misconduct, and the

27

impact of allegations in the Erhart Complaint not attributable to the alleged actionable

28

misstatements. Defendants supply a report from their own expert, David C. Smith, who
12
3:15-cv-02324-GPC-KSC

Case 3:15-cv-02324-GPC-KSC Document 247 Filed 08/24/21 PageID.5879 Page 13 of 21

1

opines that Mr. Torchio’s model is not capable of measuring only damages attributable to

2

the alleged misstatements and purports to employ methodologies that are inappropriate

3

for the case at bar. See ECF No. 211-2 (Report of David C. Smith or “Smith Report”).

4

Plaintiff responds that Defendants essentially seek to require Plaintiff to complete the loss

5

causation analysis at the class certification stage, which the Supreme Court held is not

6

required under Halliburton I. Plaintiff also provides a rebuttal report by Mr. Torchio,

7

ECF No. 226-3, which disputes Dr. Smith’s appraisal of the proposed damages model.

8

See Torchio Reply Report.
Despite Comcast’s seemingly broad pronouncements regarding the importance of

9
10

articulating a workable class-wide damages model in relation to the predominance

11

requirement, subsequent cases have generally found that this requirement does not pose a

12

significant obstacle to class certification in securities litigation. Putative class actions

13

arising from the violation of federal securities law set often forth a single theory of

14

liability—that particular material misrepresentations caused putative class members to

15

purchase the security at an artificially inflated price, which was then reduced as a result

16

of one or more corrective disclosures, resulting in damages.4 See Di Donato v. Insys

17

Therapeutics, Inc., 333 F.R.D. 427, 446 (D. Ariz. 2019). “Courts regularly reaffirm that

18

the out-of-pocket, or event study, method matches plaintiffs’ theory of liability under

19

Section 10(b) of the Securities Exchange Act, making it the standard method for

20

calculating damages in virtually every Section 10(b) class action.” Id. (quoting City of

21

Miami Gen. Employees’ & Sanitation Employees’ Ret. Trust v. RH, Inc., No. 17-CV-

22

00554-YGR, 2018 WL 4931543, at *3 (N.D. Cal. Oct. 11, 2018) (collecting cases)); see

23

also In re Diamond Foods, Inc., Sec. Litig., 295 F.R.D. 240, 251 (N.D. Cal. 2013). That

24

is because an event study aims to calculate economic loss by looking to the artificial

25
26
27
28

4

Here, while Plaintiff also asserts a claim under Section 20(a), this provision merely seeks to hold
controllers of an entity liable for the underlying Section 10(b) claim and does not inject any additional
individualized issues. See 15 U.S.C. § 78t(a).
13
3:15-cv-02324-GPC-KSC

Case 3:15-cv-02324-GPC-KSC Document 247 Filed 08/24/21 PageID.5880 Page 14 of 21

1

inflation in the stock price on each day of the class period—i.e., comparing the decline in

2

stock price following the corrective disclosure to each day of the class period prior to

3

disclosure, while controlling for expected market changes and other confounding

4

factors—permitting calculation of what the share price would have been absent the fraud.

5

See Torchio Report ¶¶ 42–69, 95–96; In re Diamond Foods, 295 F.R.D. at 248–49, 251–

6

52.
While Dr. Smith questions the efficacy of what he refers to as the “backcasting”

7
8

model employed by Mr. Torchio in general,5 Defendants primarily argue that Mr.

9

Torchio’s proposed model is insufficient to calculate damages in this case because it does

10

not account for a number of factors that could have influenced the BofI share price aside

11

from the alleged corrective disclosures, and does not include specific alternative methods

12

for calculating damages in the event some or all of the allegations in the Erhart

13

Complaint are false or there is otherwise a mismatch between the corrective disclosures

14

and the alleged misrepresentations. ECF No. 211 at 19–29. However, none of these

15

arguments convince the Court that the proposed event study would be inconsistent with

16

Plaintiff’s theory of liability such that, under Comcast, the case fails to satisfy the Rule

17

23(b)(3) predominance requirement.
First, the Court agrees with Plaintiff that under the Ninth Circuit’s ruling, at this

18
19

stage, the allegations in the Erhart Complaint need not be proven true to qualify as

20

corrective disclosures; rather, they must be reasonably perceived as true by the market

21

and correct the market’s mistake as to BofI’s underwriting standards and system of

22

internal controls and compliance infrastructure. In re BofI, 977 F.3d at 792–93.

23

Defendants’ arguments in this regard are thus inapposite. Of course, for Plaintiff to

24

ultimately show loss causation, there must be some “fit” between the allegations in the

25
26
27
28

While Dr. Smith states that the “backcasting” model is not appropriate in most cases, the Court finds
Mr. Torchio’s explanation, along with the numerous district court cases approving of similar studies,
more convincing. Smith Report ¶¶ 40–44; Torchio Reply Report ¶¶ 8–27.
5

14
3:15-cv-02324-GPC-KSC

Case 3:15-cv-02324-GPC-KSC Document 247 Filed 08/24/21 PageID.5881 Page 15 of 21

1

Erhart Complaint and Defendants’ earlier misrepresentations or else the allegations will

2

fail to qualify as corrective disclosures. But the question at class certification is whether

3

“damages could feasibly and efficiently be calculated once the common liability

4

questions are adjudicated,” Levya, 716 F.3d at 514, not whether the common liability

5

questions will be resolved in Plaintiff’s favor. Other district courts have rejected similar

6

challenges as irrelevant to the predominance inquiry underlying Comcast. See Hatamian

7

v. Advanced Micro Devices, Inc., No. 14-CV-00226 YGR, 2016 WL 1042502, at *9

8

(N.D. Cal. Mar. 16, 2016) (quoting Halliburton I, 563 U.S. at 812) (“Defendants attack

9

the ‘fit’ between an alleged corrective disclosure and a prior alleged fraudulent statement.

10

This is nothing more than an attack on loss causation, or Plaintiffs’ ability to ‘show that a

11

misrepresentation that affected the integrity of the market price also caused a subsequent

12

economic loss.’”); Baker v. SeaWorld Ent., Inc., No. 14CV2129-MMA (AGS), 2017 WL

13

5885542, at *14 (S.D. Cal. Nov. 29, 2017) (same).

14

To the extent there are allegations in the Erhart Complaint that are unrelated to

15

BofI’s previous misstatements regarding underwriting standards, internal controls, and

16

compliance infrastructure, yet still may have had an independent influence on the share

17

price, the influence of these factors will have to be taken into account in any damages

18

model. But Mr. Torchio’s report explains methods for controlling for confounding

19

variables and “overcorrections,” indicating that this could be done using the event study

20

model. Torchio Report ¶¶ 59–66, 95–96; Torchio Reply Report ¶¶ 17. The proposed

21

model therefore appropriately aligns with Plaintiff’s theory of liability. As numerous

22

cases have recognized, “[a]lthough plaintiffs may face substantial hurdles in actually

23

proving loss causation and out-of-pocket damages, they are not required to make these

24

showings until the merits stage.” In re Barrick Gold Sec. Litig., 314 F.R.D. 91, 106

25

(S.D.N.Y. 2016); cf. In re Diamond Foods, 295 F.R.D. at 252 (“Whether plaintiff will

26

ultimately prevail in proving damages is not necessary to determine at this stage.”); In re

27

Snap Inc. Sec. Litig., 334 F.R.D. 209, 217 (C.D. Cal. 2019) (“To the extent that

28

Defendant’s arguments seek to challenge [plaintiff’s expert’s] ability in practice to adjust
15
3:15-cv-02324-GPC-KSC

Case 3:15-cv-02324-GPC-KSC Document 247 Filed 08/24/21 PageID.5882 Page 16 of 21

1

for the confounding factors that may prevent isolation of price movement based on the

2

alleged partial corrective disclosures, this is an inquiry to consider at the merits stage.”);

3

SEB Inv. Mgmt. AB v. Symantec Corp., 335 F.R.D. 276, 288 (N.D. Cal. 2020)

4

(“Calculating the actual inputs into the out-of-pocket method by parsing and scaling the

5

abnormal returns requires an analysis of loss causation. For present purposes, one need

6

only realize that the inflation-ribbon inputs will be common and applied classwide.”);

7

RH, Inc., 2018 WL 4931543, at *4 (finding defendant’s argument “prematurely addresses

8

the quantification and allocation of damages, which courts consistently find are not

9

appropriately raised at the class certification stage”).

10

The logic of these other cases is consistent with Comcast and the rationale behind

11

reviewing proposed damages models at the class certification stage. The issue in

12

Comcast, the majority indicated, the damages model was not even designed to account

13

for potentially major differences between putative class members depending on the

14

theory of liability, as cable subscribers in different counties may have been harmed by

15

different forms of anticompetitive conduct, so the predominance requirement was not

16

satisfied. Comcast, 569 U.S. at 36–37. In contrast, Defendants’ dispute over whether the

17

damages model will accurately measure loss caused by corrective disclosures is unrelated

18

to the concern that individualized damages issues may predominate. Rather, they

19

highlight potential problems of proof that would apply equally to all putative class

20

members’ claims. Whether this case proceeded as a class action or many individual

21

actions, each plaintiff would confront the same loss causation issue, as each plaintiff’s

22

damages would depend on how much of the market decline could be attributed to

23

Defendants’ fraudulent conduct and how much was due to other factors. Like the

24

Supreme Court has noted with respect to the issue of materiality, the class here “will

25

prevail or fail in unison” on the issue of whether their losses were caused by

26

misrepresentations revealed by corrective disclosures in the Erhart Complaint, or were

27

driven by something else; “[i]n no event will the individual circumstances of particular

28

class members bear on the inquiry.” See Amgen, 568 U.S. at 460; cf. Insys Therapeutics,
16
3:15-cv-02324-GPC-KSC

Case 3:15-cv-02324-GPC-KSC Document 247 Filed 08/24/21 PageID.5883 Page 17 of 21

1

333 F.R.D. at 447–48 (noting that defendants’ argument was not that individual questions

2

predominate, but rather that model would not isolate the decline in stock price due to the

3

misstatements at issue, and finding this was a contention more appropriate for the merits

4

stage); In re Diamond Foods, 295 F.R.D. at 252 (“[P]laintiff has sufficiently shown that

5

damages are capable of measurement on a classwide basis such that individual damage

6

calculations do not threaten to overwhelm questions common to the class.”).

7

One securities case, arising from BP’s Deepwater Horizon drilling rig explosion,

8

provides an example of how a damages model may fail the predominance requirement in

9

a securities case. See In re BP p.l.c. Sec. Litig., No. 10-MD-2185, 2014 WL 2112823

10

(S.D. Tex. May 20, 2014), aff’d sub nom. Ludlow v. BP, P.L.C., 800 F.3d 674 (5th Cir.

11

2015). There, the court refused to certify a sub-class because the damages model sought

12

not to calculate out-of-pocket damages, but rather “consequential damages”—that is, all

13

losses after the explosion, even those not directly resulting from BP’s

14

misrepresentations—because had BP revealed the truth about the riskiness of its safety

15

program to begin with, plaintiffs would have had an opportunity to completely divest or

16

refrain from investing prior to the explosion. Id. at *10–12. The court found that this

17

theory “injects individualized inquiries into what is supposed to be a classwide model of

18

recovery” because damages would depend on whether a particular investor would or

19

would not have purchased the security had BP disclosed the safety risks. Id. at *11–12.

20

Here, in contrast, Defendants do not address how the asserted shortcomings in Mr.

21

Torchio’s damages model have any bearing on the appropriateness of class treatment.

22

This is not to say that any model of damages, no matter how arbitrary, will suffice

23

merely because the plaintiff asserts the methodology could be applied class-wide. In

24

Lortiz, for instance, the district court refused to certify a class as to damages because the

25

plaintiffs had not provided any model of damages in their opening brief, and on reply

26

only provided a rebuttal report from Mr. Torchio that addressed general techniques for

27

computing damages not tied to the facts of the case. Loritz v. Exide Techs., No. 2:13-CV-

28

02607-SVW-E, 2015 WL 6790247, at *22 (C.D. Cal. July 21, 2015). Likewise, in the
17
3:15-cv-02324-GPC-KSC

Case 3:15-cv-02324-GPC-KSC Document 247 Filed 08/24/21 PageID.5884 Page 18 of 21

1

consumer class action Werdebaugh, the district court noted that it did not “conclude that

2

individual issues predominate on damages calculations because some consumers suffered

3

more or less harm than others,” but rather found Rule 23(b)(3)’s predominance

4

requirement not satisfied because the model, as designed, included erroneous

5

assumptions and did not control for confounding information that made it impossible to

6

isolate the impact of defendant’s mislabeling from other factors. Werdebaugh v. Blue

7

Diamond Growers, No. 12-CV-02724-LHK, 2014 WL 7148923, at *14 (N.D. Cal. Dec.

8

15, 2014). Even if the Court applies the approach taken in that case to the securities

9

context, the proposed model here does seek to measure the impact on price caused by the

10

corrective disclosures alone. Cf. In re Snap, 334 F.R.D. at 218 (distinguishing

11

Werdebaugh because “an event study of a company’s stock price is expressly designed to

12

filter out fluctuations in price not attributable to specific alleged disclosures in question”);

13

Torchio Report ¶¶ 61–66; Torchio Reply Report ¶ 17.

14

The Court recognizes that Mr. Torchio has not precisely identified what approach

15

he will use to control for every variable in this case, instead generally explaining the

16

techniques used in an event study to adjust for confounding information’s effect on share

17

price, such as by analyzing analyst commentary to determine if the information is

18

material; considering whether the information was already incorporated into the share

19

price; or using financial analyses to differentiate price responses to different information.

20

See Torchio Report ¶¶ 61–66, 95–96. While Mr. Torchio has not yet conducted the

21

analysis of potential confounding information, his description of the model to be used in

22

this case indicates that he will do so. Id. ¶¶ 95–96. Aside from pointing to the fact that

23

the allegations in the Erhart Complaint have not been proven true, Defendants have not

24

identified any unusual confounding variables that make the proposed methodology ill-

25

suited to calculating damages in this case. Cf. In re Diamond Seafood, 295 F.R.D. at 252.

26

Likewise, the Ninth Circuit’s opinion indicates that there is no basis for treating the

27

Erhart Complaint as categorically different from other corrective disclosures, which

28

“short of an admission by the defendant or a formal finding of fraud . . . will necessarily
18
3:15-cv-02324-GPC-KSC

Case 3:15-cv-02324-GPC-KSC Document 247 Filed 08/24/21 PageID.5885 Page 19 of 21

1

take the form of contestable allegations of wrongdoing,” such that the proposed model

2

would be inappropriate. See In re BofI, 977 F.3d at 792. Of course, the Court does not

3

discount the possibility that Mr. Torchio’s analysis may ultimately be unable to

4

demonstrate that a corrective disclosure of Defendants’ misstatements—as opposed to

5

other confounding variables—caused the share price decline in this case. But to the

6

extent that Defendants argue Plaintiff must demonstrate that the model will be successful

7

at eliminating the effect of other variables on the share price at the class certification

8

stage, this would be no different than requiring proof of loss causation.

9

Ultimately, Defendants’ arguments cannot be squared with Halliburton I’s holding

10

that proof of loss causation is not required at the class certification stage or the

11

predominance standard itself. The purpose of analyzing the damages model at the class

12

certification stage is to determine whether individualized damages questions predominate

13

over other common issues, not to permit the defendant a preemptive strike on the merits.

14

Plaintiff has demonstrated a workable damages model that, should loss causation be

15

established, can feasibly calculate damages on a class-wide basis.

16
17
18
19

The Court therefore concludes that issues capable of class-wide resolution
predominate over individual issues.
2. Superiority
For a class to be certified under Rule 23(b)(3), the Court must also consider

20

whether “a class action is superior to other available methods for fairly and efficiently

21

adjudicating the controversy.” Fed. R. Civ. P. 23(b)(3). “A consideration of these factors

22

requires the court to focus on the efficiency and economy elements of the class action so

23

that cases allowed under subdivision (b)(3) are those that can be adjudicated most

24

profitably on a representative basis.” Zinser v. Accufix Rsch. Inst., Inc., 253 F.3d 1180,

25

1190 (9th Cir.), opinion amended on denial of reh’g, 273 F.3d 1266 (9th Cir. 2001)

26

(quoting 7A Charles Alan Wright, Arthur R. Miller & Mary Kay Kane, Federal Practice

27

and Procedure § 1780 (2d ed. 1986)). The Rule’s nonexhaustive list of factors include:

28

“(A) the class members’ interests in individually controlling the prosecution or defense of
19
3:15-cv-02324-GPC-KSC

Case 3:15-cv-02324-GPC-KSC Document 247 Filed 08/24/21 PageID.5886 Page 20 of 21

1

separate actions; (B) the extent and nature of any litigation concerning the controversy

2

already begun by or against class members; (C) the desirability or undesirability of

3

concentrating the litigation of the claims in the particular forum; and (D) the likely

4

difficulties in managing a class action.” Id.; see also Amchem Prod., Inc. v. Windsor, 521

5

U.S. 591, 616 (1997).

6

Plaintiff asserts that this case satisfies all of these criteria. ECF No. 205-1 at 23.

7

Defendants do not contest superiority. The Court agrees that adjudicating this case as a

8

class action would be superior to requiring class members to proceed on individual

9

claims. As noted above, many of the elements required to prove a violation of federal

10

securities law will be capable of proof on a class-wide basis and thus, concentrating

11

potential plaintiffs’ claims into a single action will promote judicial efficiency. See In re

12

Juniper Networks, 264 F.R.D. at 592 (“Where thousands of identical complaints would

13

have to be filed, it is superior to concentrate claims through a class action in a single

14

forum.”). As in many private securities actions, the class may include smaller investors

15

that would not have an interest in controlling their own litigation, or ability to do so. Cf.

16

In re Micron Techs., Inc. Sec. Litig., 247 F.R.D. 627, 635 (D. Idaho 2007) (noting that

17

“[m]any individual investors would not have the resources to pursue litigation). HMEPS

18

has been pursuing this case for several years and, as noted above, has demonstrated the

19

ability to competently manage the litigation during its time as Lead Plaintiff. And as the

20

primary individualized issues in this case are relatively minor—namely, the timing and

21

type of transaction undertaken by each investor—allowing this case to proceed as a class

22

action will not be unmanageable.

23

The Court therefore finds that class resolution is the superior method of

24

adjudication under Rule 23(b)(3). As Plaintiff has met the requirements of both Rule

25

23(a) and Rule 23(b)(3), the Court finds the class can be certified.

26
27
28

IV.

Class Definition

Lastly, the Court addresses an issue with respect to the appropriate definition of the
Class Period. Plaintiff contends that the Class Period should run from September 4,
20
3:15-cv-02324-GPC-KSC

Case 3:15-cv-02324-GPC-KSC Document 247 Filed 08/24/21 PageID.5887 Page 21 of 21

1

2013, the date of the first alleged misstatement, through October 14, 2015, the day after

2

the Erhart Complaint was filed. ECF No. 205-1 at 6; TAC ¶ 136; ECF No. 211-1

3

(Erhart Complaint); see also Erhart v. Bofi Holding Inc., 3:15-cv-02287-BAS-NLS, ECF

4

No. 1 (Oct. 13, 2015). Defendants argue that the October 14, 2015 end date is improper

5

because the Erhart Complaint received immediate press coverage and thus those who

6

entered into transactions on October 14, 2015 may have had knowledge of the

7

disclosures. ECF No. 211 at 9 n.1. Mr. Torchio also notes in his reply report that “it is a

8

convention in damages analysis that if the corrective information is disclosed before the

9

market opens, then investors who purchased on that day after the disclosure are not

10

damaged.” Torchio Reply Report ¶ 55 n.47. Accordingly, the Court finds the appropriate

11

Class Period runs from September 4, 2013 through October 13, 2015.

12
13

V.

Conclusion

For the reasons set forth above, the Court hereby:

14

1. GRANTS Plaintiff’s motion for class certification;

15

2. GRANTS Plaintiff’s motion to appoint Lead Plaintiff Houston Municipal

16
17

Employees Pension System as class representative; and
3. GRANTS Plaintiff’s motion to appoint Lieff Cabraser Heimann & Bernstein, LLP

18

as class counsel.

19

IT IS SO ORDERED.

20

Dated: August 23, 2021

21
22
23
24
25
26
27
28
21
3:15-cv-02324-GPC-KSC