Kaltman v. Petroleo Brasileiro S.A.- Petrobras

Filing 428

OPINION AND ORDER re: 255 MOTION to Appoint Counsel Pomerantz LLP as Class Counsel, MOTION to Certify Class and Appoint Class Representatives filed by Universities Superannuation Scheme Limited, Employees' Retireme nt System of the State of Hawaii, Union Asset Management Holding AG, North Carolina Department of State Treasurer. Having now fully reviewed the parties' submissions and evidence, the Court grants plaintiffs' motion for class certific ation, certifies a Securities Act Class and an Exchange Act Class, appoints North Carolina and Hawaii as class representatives for the Securities Act Class and USS as class representative for the Exchange Act Class, and appoints Pomerantz as Class Counsel for both Classes. The Clerk of Court is directed to close documents numbered 255 on the docket of this case. (As further set forth in this Order.) (Signed by Judge Jed S. Rakoff on 2/1/2016) (kko)

UNITED STATES DISTRICT COURT
SOUTHERN DISTRICT OF NEW YORK
-------------------------------------x
14-cv-9662
In re: PETROBRAS SECURITIES
LITIGATION

(JSR)

OPINION AND ORDER

-------------------------------------x
JED S. RAKOFF, U.S.D.J.
Lead Plaintiff Universities Superannuation Scheme Ltd.
("USS") brings this putative class action against Brazilian oil
company Petr6leo Brasileiro S.A. - Petrobras
of Petrobras' wholly-owned subsidiaries,
Finance, B.V.

("PGF")

1

("Petrobras"); two

Petrobras Global

and Petrobras America,

Inc.

("PAI")

various former officers and directors of Petrobras and its
subsidiaries

(the "Individual Defendants") ; 2 Petrobras'

On February 12, 2014, PGF acquired the outstanding shares of
another wholly-owned subsidiary of Petrobras, Petrobras
International Finance Company S.A. ("PifCo").
1

2

Specifically, the Individual Defendants include former
Petrobras Chief Executive Officer ("CEO") Maria das Gracas Silva
Foster, another former Petrobras CEO Jose Sergio Gabrielli, and
various other current or former executives of Petrobras or
associated companies, namely, Petrobras Chief Financial Officer
("CFO") Almir Guilherme Barbassa, Petrobras director Paulo
Roberto Costa, Petrobras director Jose Carlos Cosenza, Petrobras
director Renato de Souza Duque, Petrobras director Guillherme de
Oliveira Estrella, Petrobras director Jose Miranda Formigli
Filho, Petrobras director Silvio Sinedino Pinheiro, PifCo
Chairman and CEO Daniel Lima de Oliveira, PifCo director Jose
Raimundo Brandao Pereira, PifCo CFO Servio Tulio da Rosa Tinoco,
PifCo Chief Accounting Officer Paulo Jose Alves, PGF CEO and
"Managing Director A" Gustavo Tardin Barbosa, PGF CFO and
"Managing Director B" Alexandre Quintao Fernandes, PGF "Managing
Director A" Marcos Antonio Zacarias, PGF "Managing Director B"
1

independent auditor, PricewaterhouseCoopers Auditores
Independentes

("PwC"); and the various underwriters of

Petrobras's debt offerings

(the "Underwriter Defendants")

3

Plaintiffs allege that Petrobras was at the center of a multiyear, multi-billion dollar bribery and kickback scheme,

in

connection with which defendants made false and misleading
statements in violation of the Securities Act of 1933 (the
"Securities Act") and the Securities Exchange Act of 1934

(the

"Exchange Act").
The general details of this case are set forth in the
Court's Opinion dated July 30, 2015,

familiarity with which is

here presumed. See Opinion dated July 30, 2015, at 2-14, ECF No.
194. Plaintiffs now move to certify two classes, one for their
Securities Act claims and one for their Exchange Act claims.
Plaintiffs propose the following Class for their Securities Act
claims

(the "Securities Act Class") :

As to claims under Sections 11, 12(a) (2), and 15 of
the Securities Act of 1933, all purchasers who
purchased or otherwise acquired debt securities issued
Cornelis Franciscus Jozef Looman, and authorized Petrobras
United States Representative Theodore Marshall Helms.
3

Specifically, the Underwriter Defendants are: BB Securities
Ltd., Citigroup Global Markets Inc., J.P. Morgan Securities LLC,
Itau BBA USA Securities, Inc., Morgan Stanley & Co. LLC, HSBC
Securities (USA) Inc., Mitsubishi UFJ Securities (USA), Inc.,
Merrill Lynch, Pierce, Fenner & Smith Incorporated, Standard
Chartered Bank, Bank of China (Hong Kong) Limited, Banco
Bradesco BBI S.A., Banca IMI S.p.A., and Scotia Capital (USA)
Inc.
2

by Petrobras, Petrobras International Finance Company
S.A. ("PifCo"), and/or Petrobras Global Finance B.V.
("PGF") directly in, pursuant and/or traceable to a
May

l~,

2013 public orrering registered in the united

States and/or a March 11, 2014 public offering
registered in the United States. Excluded from the
Class are Defendants, current or former officers and
directors of Petrobras, members of their immediate
families and their legal representatives, heirs,
successors or assigns, and any entity in which
Defendants have or had a controlling interest.
Plaintiffs' Memorandum of Law in Support of Motion for Class
Certification at 1, ECF No. 256. Plaintiffs propose the
following Class for their Exchange Act claims

(the "Exchange Act

Class"):

As to claims under Sections 10 (b) and 20 (a) of the
Exchange Act of 1934, all purchasers who, between
January 22, 2010 and July 28, 2015, inclusive (the "Class
Period") purchased or otherwise acquired the securities
of Petroleo Brasileiro S. A.
( "Petrobras") , including
debt
securities
issued by
Petrobras
International
Finance Company S. A. ("Pi fCo") and/ or Petrobras Global
Finance B.V. ("PGF") on the New York Stock Exchange (the
"NYSE") or pursuant to other domestic transactions, and
were damaged thereby. Excluded from the Class are
Defendants, current or former officers and directors of
Petrobras, members of their immediate families and their
legal representatives, heirs, successors or assigns, and
any entity in which Defendants have or had a controlling
interest.

Id. Plaintiffs move to appoint four plaintiffs -- namely USS,
North Carolina Department of State Treasurer

("North Carolina"),

Employees' Retirement System of the State of Hawaii
("Hawaii") ,and Union Asset Management Holding AG ("Union")
class representatives for the Securities Act Class, and one
3

-- as

plaintiff, USS, as class representative for the Exchange Act
Class. Plaintiffs also move to appoint Pomerantz LLP
("Pomerantz") as Class Counsel for both Classes.
Defendants oppose plaintiffs' class certification motion,
arguing that plaintiffs have failed to satisfy the requirements
of Rules 23 (a)

and 23 (b) (3). The Court received briefing from

the parties and held an evidentiary hearing on December 21,
2015. At the hearing,

the Court heard the testimony of competing

expert witnesses: Dr. Steven Feinstein ("Feinstein")
plaintiffs and Dr.

Paul Gompers

Transcript dated Dec. 21,

("Gompers")

2015, ECF No.

for

for defendants. See

414. Each of these

experts also submitted two written reports apiece, all four of
which the Court received in evidence. See Declaration of Emma
Gilmore dated Oct. 23, 2015, Ex. A ("Feinstein Report"), ECF No.
264-1; Declaration of Emma Gilmore dated Nov. 23, 2015, Ex. H
("Feinstein Rebuttal Report"), ECF No.
Jared Gerber dated Nov.

6, 2015, Ex. 27

338-8; Declaration of
("Gompers Report"), ECF

No. 294-5; Declaration of Jared Gerber dated Dec. 8, 2015, Ex. A
("Gompers Rebuttal Report"), ECF No.

355.

Having now fully reviewed the parties'

submissions and

evidence, the Court grants plaintiffs' motion for class
certification, certifies a Securities Act Class and an Exchange
Act Class, appoints North Carolina and Hawaii as class
representatives for the Securities Act Class and USS as class
4

representative for the Exchange Act Class, and appoints
Pomerantz as Class Counsel for both Classes.
To prevail on their motion for class certification,
plaintiffs must first satisfy the four requirements of Rule
23(a), commonly referred to as numerosity,

commonality,

typicality, and adequacy. See Fed. R. Civ.

P. 23(a). The Court

considers each in turn.
Rule 23 (a) (1)

provides that class may be certified only if

"the class is so numerous that joinder of all members is
impracticable." In the Second Circuit, numerosity is usually
presumed for classes larger than forty members.

See Pennsylvania

Public School Employee's Retirement System v. Morgan Stanley &
Co.,

Inc.,

772 F.3d 111, 120

(2d Cir. 2014). However,

"the

numerosity inquiry is not strictly mathematical but must take
into account the context of the particular case." Id. Relevant
factors include " ( i)
dispersion,

(iii)

judicial economy,

(ii)

geographic

the financial resources of class members,

their ability to sue separately, and

(v)

(iv)

requests for injunctive

relief that would involve future class members." Id.
Defendants do not dispute the statements in Feinstein's
report that,

on average during the Class Period, there were

756.1 million Petrobras common ADS outstanding and 741.8 million
Petrobras preferred ADS outstanding and that the total face
value of Petrobras bonds was $41.l billion.
5

Feinstein

Report~~

33,

93, 193. On the basis of these figures,

plaintiffs estimate

that there are thousands of class members, dispersed across the
globe. Defendants do not object to this assessment per se, but
argue instead that the volume of "opt-out" individual actions
filed against Petrobras demonstrates that the class includes
sophisticated members with the resources to sue separately. See,
~,

New York City Employees Retirement System et al v.

Petroleo Brasileirio S.A. -Petrobras et al, No. 15-cv-2192.
Defendants also point to the fact that the Court has scheduled a
joint trial of the instant action and the individual actions as
evidence that a class action is not necessary in this instance.
See Order dated Nov. 18, 2015, ECF No.

311

(setting common trial

date for all cases related to the present action) .
Defendants are correct that a significant volume of
sophisticated plaintiffs have opted out of the present action,
but they miss the point of these opt-outs. The Second Circuit
has made clear that "the numerosity inquiry .

. must take into

account the context of the particular case." Pennsylvania Public
School Employee's Retirement System v. Morgan Stanley & Co.,
Inc.,

772 F.3d 111, 120

(2d Cir. 2014). The context of this

particular case is that Petrobras was among the world's largest
companies during the Class Period. Defendants do not dispute
that the billions of Petrobras securities traded vigorously
around the world throughout the Class Period.
6

In light of this,

the volume of sophisticated opt-outs does not indicate that a
class action is inappropriate or that the Classes are
insufficiently numerous. Instead, the volume of opt-outs
underscores just how vast the Classes are. Hundreds of opt-outs
is a large number, but a conservative estimate would place the
size of the proposed Classes in the thousands. Judicial economy
will be served by a joint trial because of the similarities
between the individual actions and the present action, but,
contrary to defendants'

suggestion, this would not extend to a

joint trial for thousands upon thousands of individual actions.
Accordingly, the Court concludes that the Classes satisfy the
numerosity requirement of Rule 23 (a) (1).
Rule 23(a) (2)

requires that there be "questions of law or

fact common to the class." In the context of a securities class
action, "[c]ommon questions of law and fact include whether
certain statements were false and misleading, whether those
statements violated the federal securities laws, whether those
statements were knowingly and recklessly issued, and ensuing
causation issues." Pennsylvania Ave. Funds v.
2732544 at *4

Inyx Inc.,

2011 WL

(S.D.N.Y. July 5, 2011). Common questions of law

and fact in this case include the truth of the bribery and
kickback allegations against Petrobras, the accuracy of
Petrobras's statements in connection with the allegations, the
knowledge of individual defendants regarding these matters, and
7

related causation issues. Defendants do not seriously challenge
that common questions of law and fact exist here. Accordingly,
the Court concludes that the commonality requirement is
satisfied.
Rule 23(a) (3)

requires that "the claims or defenses of the

representative parties [be] typical of the claims or defenses of
the class." Defendants do not materially attack the typicality
of North Carolina's or Hawaii's Securities Act claims, but they
argue that, because the Court dismissed Union's and USS's Notes
claims, Rule 2 3 (a) ( 3) bars them from serving as class
representatives for the Securities Act Class. Plaintiffs respond
that Union's and USS's Exchange Act claims arise from the "same
set of concerns" as the Securities Act claims, and so Union's
and USS's claims are still typical of the Securities Act Class.
NECA-IBEW Health & Welfare Fund v. Goldman Sachs & Co.,

693 F.3d

145, 149 (2d Cir. 2012). However, the "same set of concerns"
standard pertains to class standing, a distinct inquiry from
typicality under Rule 23 (a) (3).

Id. at 158 n. 9.

While the underlying thrust of plaintiffs' argument might
still have some relevance to a Rule 23(a) (3)

analysis in general

-- because "a class representative can establish the requisite
typicality under Rule 23 if the defendants

'committed the same

wrongful acts in the same manner against all members of the
class.'" Hevesi v. Citigroup,

Inc.,
8

366 F.3d 70,

82

(2d Cir.

2004)

(citation omitted)

here, Union and USS fail to clear

even this relatively modest hurdle with respect to the
Securities Act Class because they no longer have Securities Act
claims. Indeed, although plaintiffs' proposed definition of the
Securities Act Class does not explicitly require that class
members have purchased Notes in domestic transactions,

such a

requirement must be part of any certified class definition. See
Morrison v. National Australia Bank Ltd.,

561 U.S. 247

(2010)

Because neither Union nor USS adequately pleaded that they
purchased Notes in domestic transactions,

see Opinion and Order

dated Dec. 21, 2015, at 12, ECF No. 374, they cannot be members
of the Securities Act Class. And, while typicality does not
require identity amongst class members' claims, it does demand
that a class representative be a member of the Class.
Accordingly, USS and Union cannot serve as class representatives
for the Securities Act Class.
Turning to the Exchange Act Class, defendants argue that
typicality also bars USS from serving as a class representative
for the Exchange Act Class because USS's Notes claims were
dismissed. But there is no dispute that USS is a member of the
Exchange Act Class, although its claims are based only on its
purchases of Petrobras equities. Defendants object that there
are significant differences, including differences in price
movements, between Petrobras's debt and equity securities. But
9

such variations are not relevant when the same alleged
misconduct drives the claims based on debt and equity alike. The
defendants allegedly "'committed the same wrongful acts in the
same manner against all members of the class'" by participating
in a bribery and kickback scheme and making false and misleading
statements that impacted all members of the Exchange Act Class.
Hevesi v. Citigroup,
(citation omitted),

Inc.,

366 F.3d 70,

82

(2d Cir. 2004)

see In re Enron Sec. Litig., 206 F.R.D. 427,

445-46 (S.D. Tex. 2002)

("[C]ourts have repeatedly concluded

that stock purchasers can represent purchasers of debt
instruments and vice versa in the same action.")

(collecting

cases). Accordingly, the Court concludes that the typicality
requirement does not bar USS from serving as class
representative for the Exchange Act Class solely because of its
lack of Notes claims.
Defendants also argue that USS fails the typicality
requirement because it faces unique defenses in four respects.
First, defendants argue that USS is atypical because USS made
some additional purchases of Petrobras securities in June 2015,
after Petrobras had made corrective disclosures and plaintiffs
had filed the Consolidated Amended Complaint in this case. But
aside from the irrelevance of post-disclosure purchases to
earlier reliance, see In re Monster Worldwide,

Inc. Sec. Litig.,

251 F.R.D. 132, 135 (S.D.N.Y. 2008), the Class Period for the
10

Exchange Act Class runs through July 28, 2015, based on
plaintiffs' allegations that Petrobras's earlier corrective
disclosures were a "whitewash." See Opinion and Order dated Dec.
21, 2015, at 12-14, ECF No.

374.

4

Accordingly, USS's purchases of

securities in June 2015 do not mean it will face atypical
defenses.
Second, defendants argue that USS is atypical in that it
alternated between purchases and sales throughout the class
period. But such "in-and-out" trading is not atypical in a class
that contains, by defendants'

own admission, numerous

sophisticated institutional investors. See Defendants'

Joint

Memorandum of Law in Opposition to Plaintiffs' Motion for Class
Certification ("Def. Opp.") at 3, ECF No. 295. Moreover,
plaintiffs claim that USS lost approximately $80 million, its
in-and-out trading notwithstanding.

See Class Plaintiffs' Reply

Memorandum of Law in Further Support of Motion for Class
Certification at 3, ECF No.

337.

Third, defendants claim that USS's trading decisions were
based on atypical considerations. In particular, defendants
claim that USS had special contact with Petrobras during the

In their opposition to plaintiffs' motion for class
certification, defendants reiterate their earlier request to
shorten the periods for the claims in this case. The Court again
denies this request for the reasons stated in its decision on
defendants' motion to dismiss the Fourth Amended Complaint. See
Opinion and Order dated Dec. 21, 2015, at 12-14, ECF No. 374.

4

11

Class Period that affected its decisions, and also that USS
followed a special investment strategy that "look[s] at extra
financial factors" that "the market does not accurately
reflect." Declaration of Jared Gerber dated Nov.

6, 2015, Ex.

4,

ECF No. 294. Such general statements do not seriously call the
typicality of USS's claims into question: it is common practice
for money managers to claim they have some special strategy that
will deliver insights -- and returns -- superior to the wider
market. Likewise, the interactions with Petrobras that
defendants point to -- communications with the company's
investment relations team and operating personnel and a brief
meeting with the Petrobras CEO. See Declaration of Emma Gilmore
dated Nov. 23, 2015, Ex. A (Deposition of Christopher Shale) at
84:3-23, 108:10-14, ECF No.

338-1 -- are typical of the

relationships between large institutional investors and
companies like Petrobras. In a class so heavily populated by
institutional investors, these sorts of interactions do not mean
that USS is subject to atypical defenses.
Fourth, defendants claim that USS will face unique reliance
defenses based on its May 25, 2015, vote against approving
Petrobras's management reports and financial statements for
2014. These documents were part of Petrobras's alleged
"whitewash" of the bribery and kickback scandal, which valued
the total overcharges from the bribery scheme at $2.5 billion.
12

See Fourth Amended Complaint

~~

169, 176 ECF No.

342; Def. Opp.

at 7. USS objected to the documents because it had "concerns
regarding the reliability of the reported numbers." Declaration
of Jared Gerber dated Nov.

6, 2015, Ex. 3, ECF No. 294. Such

statements by plaintiffs may form part of a reliance defense,
but any such defense will be typical of the Exchange Act Class
because the Consolidated Amended Complaint in this case was
filed on March 27,

2015, and alleged that the bribery scheme

cost an estimated $28 billion. Consolidated Amended Complaint

~

5, ECF No. 109. Indeed, defendants have already argued that
plaintiffs cannot prove reliance on Petrobras's May 25, 2015,
statements because of the filings in this case. See Defendants'
Memorandum of Law in Support of their Motion to Dismiss the
Third Consolidated Amended Complaint at 10-14, ECF No. 226. The
Court takes no position on the merits of this issue at this
stage, but it does conclude that disputes over class members'
reliance on the alleged "whitewash" are typical of the Exchange
Act class a whole. Accordingly, USS's claims and defenses
against them are typical, and plaintiffs have satisfied the
requirements of Rule 23(a) (3).
Rule 23(a) (4)

requires that "the representative parties

will fairly and adequately protect the interests of the class."
"Adequacy 'entails inquiry as to whether: 1) plaintiff's
interests are antagonistic to the interest of other members of
13

the class and 2) plaintiff's attorneys are qualified,
experienced and able to conduct the litigation.'" In re Flag
Telecom Holdings, Ltd. Sec. Litig., 574 F.3d 29,

35

(2d Cir.

2009). Defendants argue that USS's interests are antagonistic to
members of the Exchange Act Class whose claims are based on
purchases of Notes or preferred ADS because USS no longer has
Notes claims and sold its preferred ADS in October 2013. See
Declaration of Jared Gerber dated Nov.

6, 2015, Ex. 16 at 1-2,

ECF No. 294-2. However, even assuming that the date USS sold its
preferred ADS would significantly alter its interests with
respect to those securities, defendants have not sufficiently
explained why the interests of holders of common ADS like USS
would be antagonistic to the interests of holder of Notes or
preferred ADSs. The only theory of antagonism of which the Court
is aware was presented during consideration of appointment of
Lead Plaintiff and concerned the differing priority of
securities in the event of bankruptcy. See Memorandum dated May
18, 2015, at 10 n.3, ECF No. 166. There is no evidence that the
bankruptcy scenario is remotely likely or relevant. Because the
same alleged misconduct drives plaintiffs' claims, regardless of
whether they arise from purchases of Notes, common ADS, or
preferred ADS, the interests of all members of the Exchange Act
Class are aligned. See In re Enron Sec. Litig., 206 F.R.D. 427,
445-46

(S.D. Tex. 2002)

("[C]ourts have repeatedly concluded
14

that stock purchasers can represent purchasers of debt
instruments and vice versa in the same action.")

(collecting

cases). Moreover, the solution to USS's putative adequacy
problem would not be to deny certification of the Exchange Act
Class but rather to appoint another class representative
alongside USS. For now, this course remains a solution in search
of a problem. However,

if, as the litigation proceeds, an

Exchange Act Class member with claims based on Notes or
preferred ADS purchases wishes to appoint a class representative
dedicated to their interests, the Court will entertain her
motion.
Defendants also argue that North Carolina, Hawaii, and USS
are collectively inadequate class representatives because they
suffer from a lack of cohesion. In particular, they rely on this
Court's decision appointing USS Lead Plaintiff to criticize the
appointment of three class representatives who, defendants
claim, are an "artificial grouping" and will not be able to
cooperate effectively. See Memorandum dated May 18, 2015, at 4
ECF No. 166. Although the Court recognizes that there are costs
associated with the appointment of multiple class
representatives, the dangers are not the same as those presented
by lawyers bundling unrelated clients together to win a lead
plaintiff appointment under 15 U.S.C.
4

(a)

(3) (B)

(iii) (I) (bb)

§

78u-

(directing court to adopt presumption
15

that lead plaintiff is person or group of persons with "the
largest financial interest in the relief sought by the class")
There is now a valid reason to appoint multiple class
representatives because Lead Plaintiff USS is no longer a member
of one of the Classes to be certified. Moreover,

the proposed

class representatives have already demonstrated that they can
work together effectively: they managed the addition of three
named plaintiffs, Hawaii, North Carolina, and Union,

and

produced a Joint Prosecution Agreement. See Order dated March
30, 2015, ECF No. 112. In light of this, the number of class
representatives is not a barrier to their collective adequacy.
Defendants make other attacks on the competence and
qualifications of the proposed class representatives and their
counsel, but none has merit.

First, defendants argue that USS

has never led a U.S. securities class action before. However,
experience is not a prerequisite to adequacy under Rule
23 (a) ( 4) .
Second, defendants claim that the volume of opt-outs should
be seen as a vote of no confidence in USS's leadership of the
class.

Defendants do not provide any support for this

interpretation of class members exercising their opt-out rights.
Indeed,

it is not uncommon for large institutions to opt out of

class actions simply so that they can improve their bargaining
position if, as usually occurs,

settlement discussions begin. If
16

anything, as explained above,

the Court views the volume of opt-

out plaintiffs as indirect evidence that a class action is
appropriate in this case and that a sophisticated institutional
investor like USS is needed as a class representative for the
thousands of remaining class members.
Third, defendants claim that the proposed class
representatives have exhibited "stark discovery failures." Def.
Opp. at 9. But this Court almost never refers discovery disputes
to Magistrate Judges, precisely so that the Court can remain
apprised of any discovery defalcations, and to this end, the
Court provides a mechanism for swift joint telephone conferences
to resolve any such problems.

If defendants felt that plaintiffs

and their counsel were behaving so badly, they should have
notified the Court sooner than their opposition to plaintiffs'
motion for class certification. The argument thus smacks more of
strategy than substance.
In any event,

on the basis not only of USS's counsel's

prior experience but also the Court's observation of its
advocacy over the many months since it was appointed lead
counsel, the Court concludes that Pomerantz, the proposed class
counsel, is "qualified, experienced and able to conduct the
litigation." In re Flag Telecom Holdings, Ltd. Sec. Litig.,
F.3d 29,

35

574

(2d Cir. 2009). There is no real dispute that

Pomerantz is an established firm with considerable class action
17

experience, and the Court has now had multiple opportunities to
observe Pomerantz's performance. The Court finds that the
Pomerantz firm has both the skill and resources to represent the
Classes adequately.
On the basis of the foregoing discussion, the Court
concludes that the requirements of Rule 23(a) (4)

are satisfied.

With that, the Court concludes that plaintiffs have satisfied
all four prongs of Rule 23(a).
In addition, of course,

for plaintiffs to prevail on their

motion for class certification, the action must meet one of the
three alternative conditions of Rule 23(b). Plaintiffs argue
that the requirements of Rule 23(b) (3)
23(b) (3)

are satisfied. Rule

requires that "a class action is superior to other

available methods for fairly and efficiently adjudicating the
controversy" and that "the questions of law or fact common to
class members predominate over any questions affecting only
individual members."
The foregoing analysis under Rule 23(a)

supports a finding

that a class action is superior to other methods of
adjudication. Petrobras was a massive company with investors
around the globe. Notwithstanding Petrobras's size and its
numerous and far-flung investors, the interests of the class
members are aligned and the same alleged misconduct underlies
their claims. Moreover, the thousands of individual class
18

members who have not opted-out have a minimal interest in
controlling the course of the litigation; there are significant
efficiency gains to be reaped from concentrating the litigation
in a single forum; and the likely difficulties in managing the
class action are readily surmountable. See Fed. R. Civ. P.
23(b) (3) (A),

(C),

(D). Defendants again point to the volume of

actions brought by individual plaintiffs as evidence against the
superiority of the class action form in this case. See Fed. R.
Civ. P. 23 (b) (3) (B). But the Court again disagrees: instead, the
volume of opt-outs demonstrates the need for a class action in
these circumstances. Otherwise, the Court risks the present
stream of individual actions growing into an unmanageable flood.
Defendants raise two more specific arguments against the
superiority of a class action in this case. First, defendants
argue that plaintiffs must demonstrate "a probability that a
foreign court will recognize the res judicata effect of a U.S.
class action judgment" to satisfy superiority. In re Vivendi
Universal, S.A., 242 F.R.D. 76,

95

(S.D.N.Y. 2007). The Court is

not aware of any binding precedent that sets out such a
requirement.

In re Vivendi Universal, S.A., the case on which

defendants rely for their position, was decided before Morrison
v. National Australia Bank Ltd.,

561 U.S. 247

(2010), which

limited the reach of U.S. securities laws to securities traded
on a U.S. exchange or purchased in domestic transactions.
19

Morrison,

561 U.S. at 267. Morrison materially lessens the

foreign res judicata concerns animating In re Vivendi Universal,
S.A .. Moreover,

In re Vivendi Universal, S.A. only concluded

that res judicata concerns could be one consideration that could
lead to the exclusion of foreign members from a class.
Vivendi Universal, S.A., 242 F.R.D. 76,

95

In re

(S.D.N.Y. 2007).

While defendants also propose including in the Class definitions
lists of countries whose residents would be excluded from the
Classes, see Defendants'

Joint Supplemental Memorandum of Law in

Further Opposition to Plaintiffs' Motion for Class
Certification, App. A, ECF No. 389, defendants have not
explained in any detail why these particular countries would not
recognize a U.S. class action judgment in this case.
Accordingly, the Court concludes that foreign res judicata
concerns are not a bar to the superiority of a class action and
declines to list any specific countries in the Class
definitions.
Defendants also argue against superiority on so-called
"ascertainability" grounds. The Second Circuit has framed
ascertainability as a stand-alone "implied requirement" of Rule
23, and, to the extent defendants' arguments are addressed to
ascertainability as distinct from superiority, the Court also
considers them here. See Brecher v. Republic of Argentina, 806
F.3d 22, 24

(2d Cir. 2015). "[T]he touchstone of
20

ascertainability is whether the class is

'sufficiently definite

so that it is administratively feasible for the court to
determine whether a particular individual is a member.'" Id.
However, "failure to certify an action under Rule 23(b) (3)

on

the sole ground that it would be unmanageable is disfavored and
'should be the exception rather than the rule.'" In re Visa
Check/MasterMoney Antitrust Litigation, 280 F.3d 124, 140

(2d

Cir. 2001).
Defendants point out that any putative class member must be
able to show that they purchased Petrobras securities on an
American exchange or in a domestic transaction under Morrison v.
National Australia Bank Ltd.,

561 U.S. 247

(2010).

Defendants

argue that, because of the nuances of the "domestic transaction"
standard, determining who is a class member and damages will be
an administratively unfeasible task for this Court,

for putative

class members who receive notice of the action, and for future
courts facing claims from class members who have not properly
opted out.

5

To cut this supposed Gordian knot, defendants propose

that the Exchange Act Class definition be amended to exclude
off-exchange purchasers and that the Securities Act Class

s Defendants also argue that the Classes are unmanageable because

plaintiffs will need to provide notice to investors across four
continents. In today's modern world, this is not an unfeasible
task, as demonstrated by the fact that Petrobras successfully
marketed its securities across four continents.
21

definition be rejected outright or amended to exclude
aftermarket purchasers and purchasers from non-U.S.
underwriters.
Amending the Class definitions in this way would cut off
purchasers who have valid claims under Morrison's second prong,
which holds that the securities laws apply to securities
purchased in "domestic transactions." Morrison,

561 U.S. at 267.

This would not be a faithful application of Morrison. Moreover,
having recently evaluated whether the four proposed class
representatives adequately pleaded that they purchased Petrobras
securities in domestic transactions, see Opinion and Order at 56, ECF No. 374, the Court is confident that the Morrison
determination is "administratively feasible." Brecher v.
Republic of Argentina,

806 F.3d 22, 24

(2d Cir. 2015). Indeed,

defendants themselves have elsewhere represented as much to the
Court. See Defendants' Supplemental Memorandum of Law in Support
of their Motion to Dismiss the Fourth Consolidated Amended
Complaint and in Further Support of their Motion to Dismiss the
Third Consolidated Amended Complaint at 6, ECF No. 351

("Each of

[Absolute Activist's tests] establishes, as the site of the
transaction that is of congressional concern, a single location
that-although subject to proof-can be easily determined based on
recognized and readily understood standards."). The criteria
identified by Absolute Activist Value Master Fund Ltd. v.
22

Ficeto,

677 F.3d 60

(2d Cir. 2012), as relevant to the

determination of whether a transaction was domestic, are highly
likely to be documented in a form susceptible to the
bureaucratic processes of determining who belongs to a Class.
For example, documentation of "the placement of purchase orders"
is the sort of discrete, objective record routinely produced by
the modern financial system that a court, a putative class
member,

or a claims administrator can use to determine whether a

claim satisfies Morrison. Accordingly, the Court concludes that
the proposed Classes are ascertainable and administratively
manageable and that a class action is the superior method of
adjudication under Rule 23 (b) (3).
Rule 23 (b) (3) also requires that "the questions of law or
fact common to class members predominate over any questions
affecting only individual members." "Class-wide issues
predominate if resolution of some of the legal or factual
questions that qualify each class member's case as a genuine
controversy can be achieved through generalized proof, and if
these particular issues are more substantial than the issues
subject only to individualized proof." UFCW Local 1776 v. Eli
Lilly and Co.,

620 F.3d 121, 131

(2d Cir. 2010). Here,

plaintiffs submit, and defendants do not meaningfully contest,
that, with the exception of reliance and damages, all elements
of plaintiffs' claims are susceptible to generalized proof. The
23

Court agrees: with the exception of reliance and damages,
plaintiffs' claims rest almost exclusively on class-wide
questions of law and fact centered around the alleged bribery
and kickback scheme, Petrobras's alleged misstatements in
connection with the scheme, the conduct of Petrobras's officers
and employees, and the effects of these actions and events on
the market.
It is true that, with respect to the Exchange Act Class,
reliance is an element of plaintiffs' claims. But while reliance
may be an individual phenomenon, here plaintiffs argue that
reliance will be established on a common basis under a "fraudon-the-market" theory. See Basic, Inc. v. Levinson,
224, 241-42

485 U.S.

(1988). "[T] o invoke the Basic presumption, a

plaintiff must prove that

(3) the [security] traded in an

efficient market." Halliburton Co. v. Erica P. John Fund,
134 S. Ct. 2398, 2413

Inc.,

(2014). The Second Circuit has not adopted

a test for the market efficiency of stocks or bonds. See
Teamsters Local 445 Freight Div. Pension Fund v. Bombardier,
Inc., 546 F.3d 196, 204 n.11

(2d Cir. 2008). However,

it has

recognized that courts generally apply a set of eight factors,
known as the "Cammer factors." Id.; see Cammer v. Bloom, 711 F.
Supp. 1264, 1286 (D.N.J. 1989)

(setting out five factors);

Krogman v. Sterritt, 202 F.R.D. 467,

478

(N.D. Tex. 2001)

(considering three additional "Cammer" factors). To address
24

these factors,

plaintiffs submitted two expert reports from

their witness Feinstein.

Feinstein also testified at an

evidentiary hearing on December 21,

2015. Defendants and their

expert do not meaningfully dispute Feinstein's conclusions with
respect to all but one of the Cammer factors

(discussed below)

The Court accepts Feinstein's testimony with respect to these
other factors and concludes that they weigh in favor of finding
that Petrobras equity and debt securities traded in efficient
markets.
The Court first considers the application of the Cammer
factors to the Petrobras equity markets. The Cammer factors are
designed for equity markets and can be applied directly to the
markets for Petrobras common and preferred ADS. The first Cammer
factor considers the average weekly trading volume during the
Class Period. Specifically, "average weekly trading of two
percent or more of the outstanding shares would justify a strong
presumption that the market for the security is an efficient
one; one percent would justify a substantial presumption."
Cammer,

711 F. Supp. at 1286

(citing Bromberg & Lowenfels,

Securities Fraud and Commodities Fraud, § 8.6

4

(Aug. 1988))

Feinstein reported that 14.1% of all common ADS and 6.61% of all
preferred ADS outstanding traded on average in a given week
during the Class Period.

Feinstein Report

25

~~

61,

171. This is

well above the 2% threshold for a "strong presumption" of
efficiency discussed in Cammer.
The second Cammer factor considers analyst coverage.
Feinstein reported that over 50 analysts covered Petrobras's
securities, inarguably a significant number. Id.

~~

66, 173.

There was also extensive news coverage of Petrobras during the
Class Period. Id.

~~

71, 176.

The third Cammer factor considers whether market makers
existed for the securities at issue. Feinstein reported that
there were at least 574 market makers for Petrobras common ADS
and 147 market makers for Petrobras preferred ADS; these market
makers included Goldman Sachs, JP Morgan, Citigroup, and Morgan
Stanley. Id.

~~

78, 181.

The fourth Cammer factor considers whether an issuer was
eligible to file a Form S-3, a simplified security registration
form that can be filed by companies that have met prior
reporting requirements. A Form F-3 is the equivalent of a Form
S-3 for foreign companies; companies are eligible to file an F-3
or an S-3 form when, among other things, they have filed
Exchange Act reports for a certain time and have a float over a
certain level. Id.

~~

81.

6

Petrobras satisfied the F-3

requirements for the duration of the Class Period, except for

"Float" refers to outstanding shares minus closely-held and
restricted shares.

6

26

when it delayed release of its financials because of the
allegations that underlie this case. Id.

~~

90,

187. Petrobras

filed an F-3 form during the Class Period on August 29, 2012.
Id.
Defendants dispute the fifth Cammer factor, which looks to
"empirical facts showing a cause and effect relationship between
unexpected corporate events or financial releases and an
immediate response in the stock price." Cammer, 711 F. Supp. at
1287. Because this factor is disputed, the Court considers it
separately below.
The sixth Cammer factor 7 considers market capitalization.
The average aggregate market value of the Petrobras common ADS
during the Class Period was $16.9 billion, greater than 90% of
publicly traded U.S. companies. Feinstein Report

~

93. The

average aggregate market value of Petrobras preferred ADS during
the Period was $15.9 billion, an amount that, on its own, would
mean Petrobras was larger than 90% of publicly traded U.S.
companies.

Id.

~

190.

The seventh Cammer factor considers the bid-ask spread for
the securities at issue. The average bid-ask spread for
Petrobras common ADS over the Class Period was 0.09%, and the

7

Really the first "Krogman" factor. As noted above, in Krogman
v. Sterritt, 202 F.R.D. 467 (N.D. Tex. 2001), the court
supplemented and elaborated on the Cammer factors.
27

average bid-ask spread for Petrobras preferred ADRs was 0.08%.
Id.

~~

99, 196. By comparison, the average bid-ask spread for

all stocks in the Center for Research in Security Prices
("CRSP") database was 0.59%. Id.
The eighth Cammer factor considers the issuer's float.
Feinstein reported that none of the Petrobras common ADS were
held by insiders or affiliated corporate entities. Id.
Accordingly,

~

95.

the entire $16.9 billion average aggregate value of

Petrobras common ADS was floated during the Class Period, again
placing Petrobas in the top decile of U.S. companies. The float
for the pref erred ADS varied during the Class Period, but
averaged $15.9 billion, always exceeding the minimum requirement
for F-3 eligibility. Id.

~

185.

The Court now considers the application of the Cammer
factors to the market for Petrobras debt securities. Although
the Cammer factors were not designed for debt securities,
plaintiffs argue that they are still useful in evaluating the
efficiency of a debt securities market, particularly in
conjunction with an analysis of the equities market for the same
company. See In re Enron Corp. Sec., 529 F. Supp. 2d 644,

747-48

(S.D. Tex. 2006). To analyze the Petrobras debt markets,
Feinstein omitted some Cammer factors, modified others, and
considered additional debt-specific factors. The Court agrees

28

that the modified Cammer factors provide a useful rubric to
evaluate debt markets.
The first modified Cammer factor considers the par value
and float of the debt securities.

Feinstein reported that the

aggregate par value of Petrobras Notes totaled $41.4 billion and
was larger than 90% of all market capitalizations on the NYSE,
Amex,
246.

and NASDAQ during the Class Period.

Feinstein Report

~

Feinstein reported that no substantial portion of Petrobras

Notes was held by insiders,
the aggregate par value.

Id.

so that the float was equivalent to
~

248.

The second modified Cammer factor considers analyst and
credit rating agency coverage of the debt securities. As noted
above,

Feinstein reported that over 50 analysts cover

Petrobras's securities,

inarguably a significant number.

Id.

66,

173. There was also extensive news coverage of Petrobras.

Id.

~~

~~

71,

176. During the Class Period,

by the major credit rating agencies,
Standard & Poor's.

Id.

~~

Petrobras was covered

Fitch, Moody's,

and

231-35.

The third modified Cammer factor considers the market
makers and underwriters for the debt securities.

Feinstein

reported that there were at least 20 underwriters of the
Petrobras Bonds,
Id.

~

241.

including large and prominent investment banks.

Feinstein also opined that underwriters generally

serve as market makers for securities and that many investment
29

banks that published analyst reports covering the bonds also
served as market makers.

Id.

~

242-43.

The fourth modified Cammer factor considers institutional
ownership of the debt securities. Feinstein reported that 214
different mutual funds held one or more Petrobras bonds during
the Class Period.

Id.

~~

236-38. Feinstein opined that wide

institutional ownership indicates market efficiency because
institutional investors often conduct their own research on
securities and make investment decisions based on that research.
Id.
The fifth modified Cammer factor again considers the
ability of the issuer to file a Form S-3. As discussed above,
Petrobras satisfied the F-3 requirements for the duration of the
Class Period, except for when it delayed release of its
financials because of the allegations that underlie this case.
Id.

~

90, 187. Petrobras filed an F-3 form during the Class

Period on August 29, 2012. Id.
The sixth modified Cammer factor considers trading volume
and frequency.

Feinstein reported a table of weekly average

trading volumes for the Petrobras Notes during the Class Period.
See id. at 64 tbl.5. The volumes ranged from 1.13% to 10.95%,
with most over 2%. Id. Accordingly, all the bonds were over
Cammer's 1% threshold for a substantial presumption of
efficiency, even though the Cammer thresholds are designed for
30

common stock, which trades more frequently than bonds.
253.

Id.

~

In addition, the average number of days between successive

trades in the Notes ranged from 0.020 and 0.418 over the Class
Period.

Id.

~

257. By comparison, relatively few corporate bonds

trade more frequently than 200 days in a year.

Id.

§

255.

Feinstein concluded that the trading volumes and frequencies of
the Notes were significantly high.
The final modified Cammer factor is the fifth unmodified
Cammer factor: empirical evidence of a cause and effect
relationship between events and an immediate response in the
price of the debt securities. Because this factor is also
disputed with respect to the Notes, the Court considers it
separately below.
Defendants and their expert Gompers do not directly dispute
Feinstein's application of the foregoing Cammer factors,
unmodified or modified.

Instead, Gompers testified that the

foregoing factors are "structural factors that are necessary for
efficient markets," but not, on their own,
Report~

27. According to Gompers,

sufficient. Gompers

the fifth Cammer factor,

"empirical facts showing a cause and effect relationship between
unexpected corporate events or financial releases and an
immediate response in the stock price," is the only factor
sufficient to show market efficiency.

31

Although the Second Circuit has recognized that evidence of
causality has been considered the most important Cammer factor,
it has not held that direct evidence is always necessary. See
Teamsters Local 445 Freight Div. Pension Fund v. Bombardier,
Inc., 546 F.3d 196, 204 n.11, 207-08

(2d Cir. 2008). While some

language in Cammer supports Gompers' view that direct evidence
is essential, see Cammer, 711 F. Supp. at 1287, this Court,
which is not bound by Cammer, does not agree that only direct
evidence is sufficient to demonstrate market efficiency in
translating material disclosures into effect on market price. As
the Supreme Court recently opined, "market efficiency is not a
yes-or-no proposition," and particularly strong indications of
market efficiency from the indirect Cammer factors can lessen
the burden to be carried by the fifth,

"direct evidence" Cammer

factor. Halliburton Co. v. Erica P. John Fund,
2398, 2414

Inc.,

134 S. Ct.

(2014). Causality is notoriously difficult to prove

with certainty, even in physics or chemistry, let alone in
market analyses, because of the large number of factors involved
and the difficulty of measuring them with precision,

separating

out their interactions, etc. Where, as here, the indirect
factors overwhelmingly describe a large and well-functioning
market for Petrobras securities, common sense suggests that the
market would materially react to material disclosures.

Put

simply, Petrobras was one of the largest and most-analyzed firms
32

in the world throughout the Class Period, and such size and
sophistication raise the likelihood of an efficient market.
In any event, though it is a somewhat involved analysis,
the Court ultimately concludes that plaintiffs have satisfied
the fifth Cammer factor. To be sure, almost every aspect was
disputed. The experts even sparred over whether any direct
evidence of the fifth factor existed. Feinstein testified that
he found direct evidence of a link between events and prices
movements in Petrobras securities. Specifically, Feinstein ran
four event studies on the Petrobras equities and two on the debt
securities. Feinstein identified three categories of event
dates:

(1) dates when Petrobras filed 6-K Forms containing the

term "corrupt*", 8 excluding dates when the terms was used only in
boilerplate language;

(2) dates when Petrobras filed any 6-K

Form; and (3) dates when Petrobras released earnings statements.
He then looked at the price movements of Petrobras securities
for a given set

(or combined multiple sets) of event dates,

using a regression analysis to strip out any price movement that
was caused by external forces,

such as moves in the wider

market. Next, he compared the proportion of event dates with
statistically significant price movements to the proportion of
non-event dates with statistically significant price movements,

s Meaning the letters "corrupt" followed by any letters or no

letters.
33

concluding that there was a statistically significant difference
in proportions for common ADS and preferred ADS and across the
Petrobras Notes. See Feinstein Report

~~

148-61, 205-21, 279-86.

In other words, there were more likely to be big price movements
on days when important Petrobras events occurred, demonstrating
the markets in Petrobras securities were responsive to new
information.
Gompers challenged both the execution and the sufficiency
of Feinstein's tests.

First, Gompers objected to Feinstein's

selection of event dates. Gompers objected that by selecting
dates uses the term "corrupt*," Feinstein ignored dates on which
allegation-related information was released to the market that
did not include the specific term "corrupt*." Gompers also
identified three additional dates with 6-Ks that included the
term corruption that he argued should not have been excluded as
boilerplate dates.

Finally, Gompers claimed that Feinstein

failed to produce evidence that the information released on
various dates across all three date sets was new.

In particular,

he contended that, because Petrobras is a Brazilian company,
some information had already been released in Brazil.
Feinstein offers some specific ripostes to these points,
but the Court does not deem it necessary to discuss them at
length here. The dispute over the inclusion of event dates is
essentially about the role of subjectivity in such analysis.
34

Gompers objects that Feinstein's choice of event dates injects
subjectivity into his analysis. However, Gompers'

suggested

improvements -- including other dates with allegation-related
information, more 6-K corruption dates, analysis of whether
information was new enough -- could also be criticized as
subjective. There is always some subjectivity in analyses of
this nature, and courts would be unable to rely on expert
testimony if they could not tolerate a modest level of
subjectivity. The Court concludes that Feinstein's selection of
event dates displays only that -- a modest level of subjectivity
and that this is not fatal to his conclusions.
Gompers next objected that Feinstein should have used the
BOVESPA index, an index of stocks on the Brazilian stock market,
instead of the CRSP Market Index in his regression analysis.
Gompers contended that the BOVESPA does a better job than the
CRSP Market Index of stripping out exogenous returns. Feinstein
responded that the BOVESPA returns are not exogenous to the
Petrobras returns because, as a result of Petrobras's size and
prominence in Brazil, the BOVESPA's movements were driven in
part by Petrobras. Moreover, Feinstein re-ran his tests using
the BOVEPSA index and concluded that using the BOVESPA in his
regression analysis would not change his overall conclusions.
See Feinstein Rebuttal Report

~

83, Exhibit-7a-7w. The Court

35

credits Feinstein's testimony and concludes that his regression
analysis is sound.
Gompers further objected that the sample sizes used in
Feinstein's tests were too small and could result in "large
standard errors, broad confidence intervals, and tests having
low power." Gompers Report
Scientific Evidence,

~

3rd ed.

84

(quoting Reference Manual on

(Washington: The National Academies

Press, 2011), 255). But Feinstein pointed out that these
properties would bias his tests against finding statistical
significance -- the danger would be false negatives not false
positives. Feinstein Rebuttal Report

~

68. Moreover,

Feinstein

performed an additional bootstrap analysis and the Fisher's
Exact Test to demonstrate that his results were robust. See
Feinstein Rebuttal Report

~~

69-70, Exhibit 8a-8b. The Court

credits Feinstein's testimony and concludes that his sample
sizes do not seriously undermine his results.
Gompers still further objected that Feinstein did not
conduct tests on the Petrobras Notes using the earnings
statement date set alone, although Feinstein did use the
earnings statement date set by itself for his analysis of the
common and preferred ADS. Gompers Report

~

72. Feinstein

responded that unless bonds are close to default they are
insensitive to earnings announcements and so the earning
statements date set by itself was not an appropriate event date
36

set for the Petrobras Notes. Feinstein Rebuttal Report

~

89.

Moreover, the results of Feinstein's regression analysis on the
Petrobras Notes showed that the fixed-rate Petrobras bonds moved
in response to market interest rates, indicating the market for
Petrobras Notes was efficient. Feinstein Report

~

288-91, Ex.

7c. Accordingly, the Court concludes that the fact that
Feinstein did not use the earnings statement date set alone in
his analysis of the Petrobras Notes does not damage his
conclusions regarding the market for Petrobras debt securities.
Gompers raised some other technical objections to
Feinstein's report.

For example, he pointed out computational

errors that Feinstein made in his initial analysis. See Gompers
Report

~

76-80. Feinstein corrected these errors in his rebuttal

report, and they did not change his conclusions. Feinstein
Rebuttal Report

~~

48-49. Upon considering the magnitude of

these errors and Gompers' other critiques of Feinstein's
execution of his methodology, the Court does not deem them
substantial enough to seriously undermine Feinstein's
credibility or his conclusions regarding the efficiency of the
markets for Petrobras securities.
Concerns about execution aside, Gompers also raised
objections to the sufficiency of Feinstein's approach. First,
Gompers objected to Feinstein's conclusions because no peerreviewed academic article has used Feinstein's methodology to
37

evaluate the efficiency of a market. Feinstein's method of
comparing the proportions of statistically significant
observations in two samples is a "z-test," essentially a version
of the more famous Student's "t-test." See Reference Manual on
Scientific Evidence, 3rd ed.

(2011), 300.

9

There is no dispute

that z-tests are commonly used and widely accepted statistical
tools. See id.; Feinstein Rebuttal
Report
3rd ed.

~

Report~

37; Gompers Rebuttal

9; see also, Reference Manual on Scientific Evidence,
(2011),

591-97

(discussing epidemiological cohort study

that compares incidence of emphysema in different populations)
Both sides refer to Feinstein's methodology as an "FDT" test
because use of z-test to evaluate market efficiency was first
proposed in a law review article by three well-known securities
econometric experts, whose combined initials were "FDT." See
Paul A. Ferrillo, Frederick C. Dunbar, and David Tabak, The
"Less Than" Efficient Capital Markets Hypothesis: Requiring More
Proof from Plaintiffs in Fraud-on-the-Market Cases, 78 St.
John's L. Rev.

Bl, 119-22 (2004). Gompers contends that, because

the article was not peer-reviewed, a z-test cannot be used to
show market efficiency. Were Feinstein using a novel or

The Reference Manual on Scientific Evidence is jointly prepared
by the Federal Judicial Center and by the National Research
Council of the National Academy of Sciences. The undersigned was
one of the four federal judges who served on the committee that
oversaw the preparation of the 3rd Edition.
9

38

questionable statistical technique, the Court would place more
weight on the absence of peer review. But it is not necessary
for every application of a commonly used statistical technique
to be peer-reviewed.

Indeed, the elegance of statistical methods

is that they can be applied to data sets of varying substantive
significance,

from rates of emphysema to transactions on modern

securities markets.

10

Because the Court is convinced that the z-

test is a well-established and sound statistical technique, the
lack of peer review does not seriously undermine Feinstein's
application of the z-test.
Next, Gompers objected to Feinstein's conclusions on the
grounds that Feinstein's z-tests failed to consider the
directionality of movements in the Petrobras market. By simply
comparing the proportions of dates with statistically
significant returns,

Feinstein's z-tests did not examine whether

10

The Court is also mystified by Gompers' claim that one of the
authors of the FDT article subsequently disavowed Feinstein's
methods. Gompers states, "[i]n fact, David Tabak (one of the
authors of the St. John's Law Review article) specifically noted
that the collective evaluation required by the FDT test rendered
the methodology 'not .
. able to fully distinguish an
efficient market from an inefficient one.'" Gompers Rebuttal
Report ~ 13 (quoting Tabak, David, "Use and Misuse of Event
Studies to Examine Market Efficiency," NERA Working Paper, April
30, 2010, 7). But Tabak's sentence is, in fact, "[t]here are
several ways that versions of the FDT methodology may not be
able to fully distinguish an efficient market from an
inefficient one." Tabak, David, "Use and Misuse of Event Studies
to Examine Market Efficiency," NERA Working Paper, April 30,
2010, 7. This manner of selective quotation does not redound to
Gompers' credit.
39

a statistically significant return on a given day was positive
or negative and,

in particular, whether the price of a security

moved up or down as expected based on the precipitating market
event. Feinstein did not dispute that his z-test methodology
alone could not test directionality. Instead, he reported the
results of a supplementary analysis examining how the prices of
common and preferred Petrobras ADS moved on earning announcement
dates. See Feinstein Rebuttal Report
conduct this analysis,

~

53, Appendix-2. To

Feinstein examined analyst reports on

earnings event dates and coded their tenor as "Positive,"
"Negative," "Mixed/Neutral," or "In Line." On dates with
statistically significant returns, he found that the price
movements in common and preferred ADS were consistent with his
assessments of the tenor of analyst coverage. Id.
Gompers and defendants objected to this analysis as
subjective and flawed. Overall, they objected to Feinstein's
categorization of the tenor of analyst coverage as dependent on
his subjective interpretation. More specifically, they claimed
that the tenor of coverage on two of the dates Feinstein labeled
"Positive," May 16, 2011, and October 28, 2013,

should have been

labeled "Mixed/Neutral." The Court agrees that these dates were
mischaracterized. See Transcript dated Dec. 21,
ECF No.

413; Gompers Rebuttal Report

~~

2015 at 44-50,

52-55. Moreover,

Feinstein did not provide the analyst reports he relied on in
40

making his coverage assessments and, other than the excerpts
listed in Appendix-2 to his rebuttal report, did not explain how
he arrived at specific tenor determinations. Therefore, it is
difficult to assess whether the two dates identified by
defendants are anomalous or indicative of wider deficiencies in
Feinstein's directionality testing. Accordingly, the Court
places only limited weight on Feinstein's directionality testing
of the Petrobras ADS.
The Court also places only limited weight on the evidence
of the directionality of the movements in the Petrobras Notes
market.

Feinstein reported that his regression analysis of the

Notes showed that they moved with his Benchmark Bond return
variable, which serves as basic confirmation of the
directionality of Notes price movements.

Feinstein Report

~~

288-91; Ex. 7c. However, Gompers identified three dates when
some Notes had statistically significant price declines while
other Notes had statistically significant price increases.
Feinstein did not address these movements. Accordingly, the
Court concludes that there is only limited evidence of
directionality in the Petrobras Notes market.
However, evidence of directionality or the degree of fit
between expected and observed moves in a market need not be
substantial to allow a finding of market efficiency. Such
evidence goes to the accuracy of the price of a security, and
41

the Supreme Court has explained that it is not the accuracy of a
price as a reflection of underlying value but instead the
sensitivity of the price to false statements that underlies the
Basic presumption. See Halliburton Co. v. Erica P. John Fund,
Inc., 134 S. Ct. 2398, 2410

(2014)

("'That the

price [of a

stock] may be inaccurate does not detract from the fact that
false statements affect it, and cause loss,'" which is 'all that
Basic requires.'")
685

(7th Cir. 2010)

(quoting Schleicher v. Wendt,

618 F.3d 679,

(alteration in original)). Defendants' own

arguments that Feinstein's tenor assessments were subjective
demonstrate the wisdom of the Supreme Court's position. Any
assessment of the tenor of analyst coverage and the expected
impact of an event on the market will be subjective. Indeed, the
analyst reports released on May 15, 2011, and May 16, 2011,
varied in their assessments of the same earnings event. See
Feinstein Rebuttal Report, Appendix-2; Transcript dated Dec. 21,
2015 at 44-50, ECF No.

413. Whether the market, upon receiving

new information, moved in the precise way analysts or experts
would expect it to move is not the key to unlocking Basie's
presumption of reliance. What is essential is evidence that,
when the market received new information,

it "generally

affect[ed]" the price. Halliburton, 134 S. Ct. at 2410. In this
case, the z-tests provide such evidence. Accordingly, the Court

42

concludes that the limited evidence of directionality is not
fatal to plaintiffs'

showing of market efficiency.

Finally, Gompers objected to the sufficiency of Feinstein's
results on the grounds that "in an efficient market, the price
of a security should always move in response to the release of
new value-relevant information that is materially different from
expectations." Gompers Rebuttal Report

~

31. Gompers allowed

that, because of potential shortcomings in a regression
analysis, "there may be instances where [an] event study does
not always show directionally consistent price movements to new
information." Id. But, he "would expect the vast majority of
days with new value-relevant information that is materially
different from expectations to have statistically significant
price movements that are directionally consistent with the
information." Id. Gompers pointed out that Feinstein's event
studies failed to show that the Petrobras markets moved in
response to events the vast majority of the time.
Feinstein responded that not every event will move a market
and that the impact of an event depends on various factors,
including, among other things, the nature of the event, whether
the information involved is truly new,

11

whether a confounding

This factor is why the Court gives little weight to Gompers'
application of Feinstein's methodology to the eighty-five
alleged corrective disclosure dates in Plaintiffs' Complaint.
Gompers Report ~~ 88-89, 92, Ex. 3. Gompers found that the

11

43

event occurs simultaneously, the magnitude of background
volatility, and how the event unfolded.
Report

~

Feinstein Rebuttal

33. In light of these complex forces,

one should not

expect to see a price movement on every news day.
The Court sides with Feinstein. The Supreme Court has
rejected Gompers' absolutist view of market efficiency by making
clear that "market efficiency is a matter of degree" and that
"Basie's presumption of reliance

. does not rest on a

'binary' view of market efficiency." Halliburton Co. v. Erica P.
John Fund,

Inc., 134 S. Ct. 2398, 2410

(2014).

In assessing

market efficiency, courts should not let the perfect become the
enemy of the good.

In this case, where the indirect Cammer

factors lay a strong foundation for a finding of efficiency, a
statistically significant showing that statistically significant
price returns are more likely to occur on event dates is
sufficient as direct evidence of market efficiency and thereby
to invoke Basie's presumption of reliance at the class
certification stage. Accordingly, plaintiffs have adequately

proportion of alleged corrective disclosure dates with
statistically significant price moves was not statistically
significantly larger than other dates during the period. Id. In
contrast to Feinstein's selection of event dates, which involved
a tolerable level of subjectivity, see supra, the alleged
corrective disclosure dates were compiled by plaintiffs as dates
when news of the alleged bribery and kickback scheme trickled
out. By design, they did not all involve new information being
presented to the market and are therefore not an appropriate
sample for a z-test. Feinstein Rebuttal Report ~ 64.
44

demonstrated that common issues of law and fact will predominate
over individual issues with respect to the reliance element of
their Exchange Act claims.
Defendants also argue that plaintiffs have failed to
satisfy Rule 23(b) (3) because they have not presented an
adequate model of classwide damages. It is "'well-established'
in [the Second Circuit] that 'the fact that damages may have to
be ascertained on an individual basis is not sufficient to
defeat class certification' under Rule 23(b) (3) ."Roach v. T.L.
Cannon Corp., 778 F.3d 401,

405

(2d Cir. 2015). However, the

Supreme Court has held that if a court does rely on a classwide
model of damages when certifying a class, the "model .

. must

actually measure damages that result from the class's asserted
theory of injury." Id. at 407; see Comcast Corp. v. Behrend, 133
S. Ct. 14 2 6 ( 2013) .
Feinstein proposed a three-step damages methodology:

(1) an

event study could determine the amount of price inflation in a
given security, as well as how much of this dissipated upon
disclosures;

(2)

an "inflation ribbon" could be constructed,

measuring the difference between the inf lated price of the
security and what it would have traded at without the alleged
misrepresentations; and (3) per shares damages could be
calculated as the difference between the inflation on the date

45

shares were purchased and the inflation on the date those same
shares were sold. Feinstein Report

~

296.

In response, Gompers divided the alleged corrective
disclosures into "numeric" and "non-numeric" disclosures.
Gompers Report

~

107-08. Numeric disclosures involved

quantitative information,

such as the amount of a write-down,

while non-numeric disclosures involved qualitative information,
such as acknowledgment of ethical breaches. Id. Gompers claimed
that numeric disclosures would categorically have no impact on
the price of Petrobras securities because prices were based on
the economic value of Petrobras's assets, specifically their
future cash flows.

Id.

~~

109-17. Gompers further claimed that

the impact of non-numeric disclosures on the prices of Petrobras
securities would be too difficult to measure because, among
other reasons, different investors would have had different
appetites for risk when investing in Petrobras and price
declines following non-numeric disclosures could have been
caused by collateral factors.
It is not necessary, however, to resolve the detailed
disputes over plaintiffs' damages model at the class
certification stage. Indeed, plaintiffs do not even have a
burden to produce a classwide damages model at this time.
"' [T]he fact that damages may have to be ascertained on an
individual basis'

[is]

simply one 'factor that [courts must]
46

consider in deciding whether issues susceptible to generalized
proof 'outweigh'

individual issues' when certifying the case as

a whole." Roach v. T.L. Cannon Corp., 778 F.3d 401,

405

(2d Cir.

2015). Nonetheless, the Court concludes that plaintiffs'
proposed damages model weighs modestly, although not
dispositively,

in favor of granting class certification.

Plaintiffs' proposed damages model is not unusual for a
securities fraud class action. The Court credits Gompers' point
that there may be serious difficulties in determining the impact
of non-numeric disclosures. But it is not clear that these
difficulties will be fatal,

and they do not mean that

plaintiffs' proposed model does not match their theory of
liability. The Court does not credit Gompers' claim that numeric
disclosures have no effect on the prices of Petrobras
securities. The Court understands Gompers' point about economic
value as a theoretical matter, but,

in practical terms,

it is

difficult for the Court to accept that, in a reasonably
efficient market, a company's stock price would not decline upon
reports that it faces billions of dollars in losses. Gompers
Report

~

108, 117. Accordingly, the Court concludes that

plaintiffs' model of classwide damages provides a modest
indication that common issues of law and fact will predominate
over individual issues under Rule 23(b) (3).

47

Based on the foregoing analysis, the Court concludes that
plaintiffs have satisfied the requirements of Rule 23(b) (3).
Because plaintiffs have satisfied the requirements of Rule 23,
the Court hereby certifies two classes. The Exchange Act Class
is defined as follows:

As to claims under Sections 10 (b) and 20 (a) of the
Securities Exchange Act of 1934, all purchasers who,
between January 22, 2010 and July 28, 2015, inclusive
(the "Class Period") purchased or otherwise acquired the
securities of Petroleo Brasileiro S. A. ( "Petrobras") ,
including
debt
securities
issued
by
Petrobras
International Finance Company S. A.
("Pi fCo")
and/ or
Petrobras Global Finance B. V. ( "PGF") on the New York
Stock Exchange
(the "NYSE")
or pursuant to other
domestic
transactions,
and
were
damaged
thereby.
Excluded from the Class are Defendants, current or
former officers and directors of Petrobras, members of
their
immediate
families
and
their
legal
representatives, heirs, successors or assigns, and any
entity in which Defendants have or had a controlling
interest.

The Securities Act Class is defined as follows:
As to claims under Sections 11 and 15 of the Securities
Act of 1933, all purchasers who purchased or otherwise
acquired debt securities issued by Petroleo Brasileiro
S.A.
("Petrobras"),
Petrobras International Finance
Company S.A. ("PifCo"), and/or Petrobras Global Finance
B.V. ("PGF"), in domestic transactions, directly in,
pursuant and/or traceable to a May 15, 2013 public
offering registered in the United States and/or a March
11, 2014 public offering registered in the United States
before
Petrobras made
generally available
to
its
security holders an earnings statement covering a period
of at least twelve months beginning after the effective
date of the offerings, and were damaged thereby. As to
claims under Sections 12 (a) (2) of the Securities Act of
1933, all purchasers who purchased or otherwise acquired
debt securities issued by Petroleo Brasileiro S.A.
48

("Petrobras"), Petrobras International Finance Company
S. A.
( "P if Co") , and/ or Petrobras Global Finance B. V.
("PGF"), in domestic transactions, directly in a May 15,
2013 public offering registered in the United States
and/ or a March 11, 2014 public off er ing registered in
the United States before Petrobras made generally
available to its security holders an earnings statement
covering a period of at least twelve months beginning
after the effective date of the offerings, and were
damaged thereby. Excluded from the Class are Defendants,
current or former officers and directors of Petrobras,
members of their immediate families and their legal
representatives, heirs, successors or assigns, and any
entity in which Defendants have or had a controlling
interest.
The Court appoints USS class representative for the Exchange Act
Class and North Carolina and Hawaii class representatives for
the Securities Act Class. The Court appoints Pomerantz LLP as
class counsel for both Classes.
The Clerk of Court is directed to close documents numbered
255 on the docket of this case.

SO ORDERED.
Dated:

New York, NY
February _l_, 2016

49