United States Court of Appeals
Fifth Circuit
In the F I L E D
United States Court of Appeals August 23, 2005
for the Fifth Circuit
_______________
Charles R. Fulbruge III
m 04-11078 Clerk
_______________
RICHARD BELL,
ON BEHALF OF HIMSELF AND ALL OTHERS SIMILARLY SITUATED; ET AL.,
Plaintiffs,
PLUTARCH, LTD., MARIO SONZONE,
Plaintiffs-Appellants,
VERSUS
ASCENDANT SOLUTIONS, INC.; NORMAN CHARNEY;
PAUL JENNINGS; CCLP, LTD., ALAN E. SALZMAN,
Defendants-Appellees.
***************
DENNIS HOFFMAN,
INDIVIDUALLY AND ON BEHALF OF HIMSELF
AND ALL OTHERS SIMILARLY SITUATED; ET AL.,
Plaintiffs,
PLUTARCH, LTD.,
Plaintiff-Appellant,
VERSUS
ASCENDANT SOLUTIONS, INC.;
NORMAN CHARNEY; PAUL JENNINGS,
Defendants-Appellees.
_________________________
Appeal from the United States District Court
for the Northern District of Texas
_________________________
Before GARWOOD, SMITH, and CLEMENT, not abuse its discretion, we affirm and remand.
Circuit Judges.
I.
JERRY E. SMITH, Circuit Judge: Ascendant, a Dallas-based firm founded in
1995, provided electronic order management
Mario Sonzone and Plutarch, Ltd., an indi- and customer service solutions to e-commerce
vidual investor and a closely-held Liberian cor- and direct marketing firms. It made an initial
poration, appeal the denial of class certification public offering (“IPO”) of five million shares
in a securities fraud suit against Ascendant of common stock on November 11, 1999. All
Solutions, Inc. (“Ascendant”), some of its the shares were purchased on a firm-commit-
former executives and directors, and a related ment basis and at a pre-set price ($8.00 per
entity. Concluding that the district court did share) by an underwriting syndicate. After
2
two days of trading on the NASDAQ National ests) who purchased Ascendant common stock
Market, both of which were marked by insig- on the open market between November 11,
nificant price declines, Ascendant common 1999 (the date of the IPO) and January 24,
stock more than tripled in price within three 2000 (the day Ascendant announced its trou-
weeks, closing at $28.00 on November 30, all bles) and who were damaged by defendants’
during the twenty-five-day post-IPO “quiet allegedly false and misleading statements in
period.” See 17 C.F.R. § 230.174(d). violation of sections 10(b) and 20(a) of the Se-
curities Exchange Act of 1934 and rule 10b-5
The good times were short-lived: On Jan- promulgated thereunder.1
uary 24, 2000, Ascendant announced that
problems with its capacity to provide the re- Ascendant responded in opposition to class
quisite software services had caused it to lose certification and, in support, submitted an ex-
three of its seven customers, including one pert report, the rub of which was that Ascen-
featured in Ascendant’s prospectus as a “Se- dant’s common stock did not trade in an effi-
lect Client Case Stud[y].” The next day, As- cient market. This being so, Ascendant main-
cendant’s stock price declined almost 30%. tained, the putative class could not invoke the
By the end of September, it had announced fraud-on-the-market theory recognized in Ba-
that it would no longer provide order fulfill- sic Inc. v. Levinson, 485 U.S. 224 (1988), and
ment and customer-service call-center opera- obtain the benefit of its class-wide presump-
tions; by May 2001, it had been de-listed from tion of reliance,2 leaving plaintiffs’ fraud claims
NASDAQ.
1
II. See 15 U.S.C. § 78j(b), 78t(a); 17 C.F.R.
§ 240.10b-5.
Litigation ensued. The district court con-
solidated five securities fraud class action com- 2
The fraud-on-the-market theory enables in-
plaints filed against Ascendant and some of its vestors who cannot satisfy the traditional require-
executives and directors and appointed lead ment of proving actual reliance on a fraudulent
plaintiffs. Plaintiffs filed an amended class representation (i.e., those investors who did not
action complaint; a motion to dismiss fol- read the documents or hear the statements alleged
lowed, which the district court granted in part to contain the fraudulent representations) neverthe-
and denied in part, winnowing-down some of less to maintain a fraud action for which reliance is
the plaintiffs’ allegations but leaving their basic an essential element. It does so by “interpreting the
theory of liability intact: Ascendant and reliance requirement to mean reliance on the integ-
various insiders had made false and misleading rity of the market price rather than reliance on the
statements in connection with the IPO challenged disclosure.” Daniel R. Fischel, Effi-
regarding the scope of Ascendant’s order cient Capital Markets, the Crash, and the Fraud
on the Market Theory, 74 CORNELL L. REV. 907,
management and customer service systems and
908 (1989).
its success in providing such systems to clients.
The central premise of the theory is that, in an
Plaintiffs then moved to certify a class, efficient capital market, the market price of a stock
based on Federal Rule of Civil Procedure reflects all public information; hence an investor
23(b)(3), consisting of all persons (except de- who purchases a stock in such a market is harmed
fendants and certain related persons and inter- if the price reflects false information as a conse-
(continued...)
3
dependent on proving individual reliance and tion.5 Plaintiffs thereafter sought, and we
thus unsuited for aggregation.3 granted, pursuant to Rule 23(f) of the Federal
Rules of Civil Procedure, an interlocutory ap-
Plaintiffs responded with an expert report peal of that denial.
of their own, which included an event study
purporting to show that Ascendant common III.
stock did, in fact, trade in an efficient market. The class certification decision rests within
But the district court, on Ascendant’s motion the sound discretion of the district court, so
to exclude under Daubert,4 excluded plaintiffs’ long as that discretion is exercised within the
expert, concluding that his event study was framework of rule 23. See Robinson v. Texas
unreliable and purposefully designed to Auto Dealers Ass’n, 387 F.3d 416, 421 (5th
support its market-efficiency conclusion. Cir. 2004); Castano, 84 F.3d at 740. Thus we
review for abuse of discretion the denial of
The court then determined that plaintiffs, class certification. See In re Monumental Life
lacking an expert, had otherwise failed to dem- Ins. Co., 365 F.3d 408, 414 (5th Cir.), cert.
onstrate that Ascendant common stock traded denied, 125 S. Ct. 277 (2004).
in an efficient market, so the putative class
could not take advantage of the presumption IV.
of class-wide reliance permitted under the Plaintiffs challenge the district court’s con-
fraud-on-the-market theory. The fraud claims clusion that they failed adequately to show that
thus would require proof of individual reliance, Ascendant common stock traded in an efficient
so the proposed class does not satisfy the market during the class period. We un-
predominance requirement of rule 23(b)(3). derstand their argument on appeal to contain
Accordingly, the court denied class certifica- two primary contentions. First, they claim
they need only plead market efficiency at the
class certification stage and that the district
court, by looking beyond the pleadings, im-
properly decided an issue going to the merits
2
(...continued) under Eisen v. Carlisle & Jacquelin, 417 U.S.
quence of a material misrepresentation. See id. at 156 (1974). Second, they attack the substance
911; Basic, 485 U.S. at 246. Accordingly, the of the court’s market efficiency determination
fraud-on-the-market theory holds “only to the on the ground that the court failed to give due
extent that markets efficiently reflect (and thus consideration to various factors relevant to
convey to investors the economic equivalent of) all market efficiency.
public information”. Asher v. Baxter Int’l Inc.,
377 F.3d 727, 731-32 (7th Cir. 2004), cert. de- A.
nied, 125 S. Ct. 1639 (2005). Plaintiffs claim they are required only to
3 plead market efficiency at the class certifica-
Cf. Castano v. Am. Tobacco Co., 84 F.3d
tion stage and that the district court, by going
734, 745 (5th Cir. 1996) (“[A] fraud class action
cannot be certified when individual reliance will be
an issue.”).
5
See Bell v. Ascendant Solutions, Inc., No.
4
See Daubert v. Merrell Dow Pharms., Inc., Civ. A. 3-01-CV-0166-N, 2004 WL 1490009
509 U.S. 579 (1993). (N.D. Tex. July 1, 2004).
4
beyond the pleadings and requiring a threshold has cogently explained in rejecting a similar
showing, improperly decided an issue going to contention,
the merits under Eisen. This betrays a mis-
reading of Eisen, which, as we explained in Eisen’s prohibition against assessing plain-
Castano, does not suggest that a court is lim- tiffs’ likelihood of success on the merits as
ited to the pleadings when deciding on class part of a Rule 23 certification does not
certification. Rather, Eisen “stand[s] for the mean that consideration of facts necessary
unremarkable proposition that the strength of to a Rule 23 determination is foreclosed
a plaintiff’s claim should not affect the certifi- merely because they are required to be
cation decision.” Castano, 84 F.3d at 744.6 proved as part of the merits. The analysis
under Rule 23 must focus on the require-
Eisen therefore offers no support for the ments of the rule, and if findings made in
view that a district court must accept, on noth- connection with those requirements overlap
ing more than pleadings, allegations of ele- findings that will have to be made on the
ments central to the propriety of class certifi- merits, such overlap is only coincidental.
cation under rule 23.7 As the Fourth Circuit The findings made for resolving a class
action certification motion serve the court
only in its determination of whether the
6
See Castano, 84 F.3d at 744. (“In Eisen, the requirements of Rule 23 have been
Court held that it was improper to make a prelimi- demonstrated.
nary inquiry into the merits of a case, determine
that the plaintiff was likely to succeed, and conse- Gariety v. Grant Thornton, LLP, 368 F.3d
quently shift the cost of providing notice to the
356, 366 (4th Cir. 2004). Thus, in the market
defendant.”) (citing Eisen, 417 U.S. at 177)).
efficiency context, “[a]lthough the court’s de-
7
Cf. Gen. Tel. Co. v. Falcon, 457 U.S. 147, termination for class certification purposes
160 (1982) (“[S]ometimes it may be necessary for may be revised (or wholly rejected) by the ul-
the court to probe behind the pleadings before timate factfinder, the court may not merely
coming to rest on the certification question.”); presume the facts in favor of an efficient mar-
Coopers & Lybrand v. Livesay, 437 U.S. 463, 469 ket.” Unger v. Amedisys, Inc., 401 F.3d 316,
(1978) (reasoning that “the class determination 323 (5th Cir. 2005).8
generally involves considerations that are ‘en-
meshed in the factual and legal issues comprising
the plaintiff’s cause of action.’”); Castano, 84 F.3d 7
(...continued)
at 744 (“A district court may certainly look past certify a class cannot be found in Rule 23 and has
the pleadings to determine whether the require- nothing to recommend it. . . . Before deciding
ments of rule 23 have been met. Going beyond the whether to allow a case to proceed as a class action
pleadings is necessary, as a court must understand . . . a judge should make whatever factual and legal
the claims, defenses, relevant facts, and applicable inquiries are necessary under Rule 23. . . . And if
substantive law in order to make a meaningful some of the considerations under Rule 23(b)(3) . .
determination of the certification issues.”) (footnote . overlap the merits . . . then the judge must make
omitted)); Szabo v. Bridgeport Machs., Inc., 249 a preliminary inquiry into the merits.”).
F.3d 672, 675-76 (7th Cir. 2001) (“The proposi-
8
tion that a district judge must accept all of the Cf. Nathenson v. Zonagen, Inc., 267 F.3d
complaint’s allegations when deciding whether to 400, 414 (5th Cir. 2001) (“Basic plainly states that
(continued...) (continued...)
5
Indeed, the suggestion that a court must ac- Without an initial demonstration of market
cept mere allegations of market efficiency is efficiency, there is no assurance that the
demonstrably at odds with Unger and, more available material information concerning
fundamentally, with a district court’s duty, the stock translates into an effect on the
rooted in the text of rule 23(b)(3), to “find[]” market price and supports a classwide pre-
that common issues predominate before certi- sumption of reliance. Absent an efficient
fying a class.9 At issue in Unger were “the market, individual reliance by each plaintiff
standards and procedures used by district must be proven, and the proposed class will
courts when considering certification of secu- fail the predominance requirement.
rities class actions dependent on the ‘fraud on
the market’ theory,” and we held that “a care- Id. at 322. Accordingly, we joined several of
ful certification inquiry is required and findings our sister circuits in applying “rigorous,
must be made based on adequate admissible though preliminary, standards of proof to the
evidence to justify class certification.” Unger, market efficiency determination,” id., and we
401 F.3d at 319. set forth various factors utilized by courts to
decide whether a stock trades in an efficient
In so doing, we stressed the critical link be- market.10
tween a threshold showing of market efficien-
cy and a district court’s duty to ensure that Plaintiffs acknowledged Unger at oral argu-
class members’ fraud claims are not predicated ment but contended it is irrelevant insofar as it
on proving individual reliance: involved a small-cap stock traded on the over-
the-counter market, whereas Ascendant was
listed on the NASDAQ National Market and
traded more heavily. Unger did involve a
8 small-cap stock traded on a less-developed
(...continued)
the presumption of reliance [under the fraud-on-
the-market theory] may be rebutted by ‘[a]ny
10
showing that severs the link between the alleged These factors include (1) the average weekly
misrepresentation and . . . the price received (or trading volume expressed as a percentage of total
paid) by the plaintiff.’”) (quoting Basic, 485 U.S. outstanding shares; (2) the number of securities an-
at 247)). alysts following and reporting on the stock; (3) the
extent to which market makers and arbitrageurs
9
See Unger, 401 F.3d at 321 (“The plain text trade in the stock; (4) the company’s eligibility to
of Rule 23 requires the court to ‘find,’ not merely file SEC registration Form S-3 (as opposed to
assume, the facts favoring class certification.”); Form S-1 or S-2); (5) the existence of empirical
Gariety, 368 F.3d at 365 (“If it were appropriate facts “showing a cause and effect relationship
for a court simply to accept the allegations of a between unexpected corporate events or financial
complaint at face value in making class action releases and an immediate response in the stock
findings, every complaint asserting the require- price”; (6) the company’s market capitalization; (7)
ments of Rule 23(a) and (b) would automatically the bid-ask spread for stock sales; and (8) float, the
lead to a certification order, frustrating the district stock’s trading volume without counting insider-
court’s responsibilities for . . . making ‘findings’ owned stock. Unger, 401 F.3d at 323 (citing
that the requirements of Rule 23 have been satis- Cammer v. Bloom, 711 F. Supp. 1264, at 1286-87
fied.”) (citing FED. R. CIV. P. 23(b)(3)) (internal (D.N.J. 1989)); Krogman v. Sterritt, 202 F.R.D.
citations omitted). 467, 477-78 (N.D. Tex. 2001)).
6
market, and these distinctions are, of course, an efficient capital market.12
relevant to the substance of the market effi-
ciency determination. But we reject the sug- In any event, if plaintiffs are as confident as
gestion that Unger exempts plaintiffs in suits they seem to be in their empirical assertion that
involving stocks traded on larger securities the market for stocks, such as Ascendant,
markets from the burden of making a prelimi- listed on NASDAQ are necessarily efficient
nary showing of market efficiency at the class during the relevant class periods, then making
certification stage. a preliminary showing of such efficiency on
competent, admissible evidence should be no
Beyond the conflict with rule 23(b)(3)’s burden at all. Accordingly, we reject plaintiffs’
requirement that a court “find[]” that common contention that the district court erred in going
issues predominate and the forceful logic of beyond the pleadings and requiring them to do
Unger itself,11 as well as the obvious problems more than just allege market efficiency to
that would attend application of such a rule satisfy the predominance requirement of rule
(e.g., which markets and how much volume 23(b)(3).13
(and for how long) would suffice to free a
putative class from the threshold proof re-
12
quirement?), we note that, as explained further See, e.g., Jonathan R. Macey & Geoffrey P.
infra, the mere fact that a stock trades on a na- Miller, Good Finance, Bad Economics: An Analy-
tional exchange does not necessarily indicate sis of the Fraud-on-the-Market Theory, 42 STAN.
that the market for that particular security is L. REV. 1059, 1060 (1990).
efficient. See Cammer, 711 F. Supp. at 1281. 13
Plaintiffs also claim that the district court
In fact, “some companies listed on national
improperly required expert testimony to prove mar-
stock exchanges are relatively unknown and ket efficiency. But plaintiffs fail to grasp the
trade there only because they met the eligibility distinction between saying something has not been
requirements. While the location of where a proven without an admissible expert report and
stock trades might be relevant, it is not saying something cannot be proven without it.
dispositive of whether the ‘current price re- The district court did not hold, as plaintiffs’ brief-
flects all available information,’” id. (footnote ing suggests, that expert testimony is required as a
omitted), which, of course, is the hallmark of matter of law to show market efficiency. Rather,
the court found the particular expert testimony
offered by plaintiffs to be unreliable, then con-
cluded that there was otherwise insufficient evi-
dence of market efficiency to permit plaintiffs to
demonstrate predominance by way of the fraud-on-
11
See Unger, 401 F.3d at 325 (“Questions of the-market theory.
market efficiency cannot be treated differently from
other preliminary certification issues.”); id. at 323 In any case, although “[t]here is no requirement
(“[T[he court may not simply presume the facts in for expert testimony on the issue of market effi-
favor of an efficient market.”); id. at 325 (“When ciency . . . many courts have considered it when
a court considers class certification based on the addressing this determination, which may often
fraud on the market theory, it must engage in benefit from statistical, economic, and mathemati-
thorough analysis, weigh the relevant factors, cal analysis.” Unger, 401 F.3d at 323 n.6. Indeed,
require both parties to justify their allegations, and though Unger admonishes district courts “not to
base its ruling on admissible evidence.”). (continued...)
7
B. price movement in response to new, company-
Thus, our focus narrows to the substance of specific information. Beyond their expert re-
their market efficiency showing; we consider port, however, plaintiffs did not provide the
whether the court abused its discretion in find- court with any analysis of these or other mar-
ing that plaintiffs failed to make a showing suf- ket efficiency factors.
ficient to avail themselves of the class-wide
presumption of reliance under the fraud-on- Plaintiffs’ motion for class certification is
the-market theory. Because we hear this ap- devoid of any arguments or evidence in sup-
peal on an interlocutory basis, however, our port of the fraud-on-the-market theory. The
review is bridled by rule 23(f). Under that brief filed in support of the motion does con-
rule, “a party may appeal only the issue of tain a section on predominance, but it, too,
class certification; no other issues may be contains no analysis of any of the market ef-
raised.” Bertulli v. Indep. Ass’n of Cont’l ficiency factors. Instead it merely assumes that
Pilots, 242 F.3d 290, 294 (5th Cir. 2001).14 the putative class is entitled, as a matter of
Consequently, as plaintiffs concede, we may right, to a presumption of reliance, noting with
not review the exclusion of their expert report, little more than a citation to Basic that “the re-
so we must look to the remainder of their mar- liance element is presumed.”
ket efficiency showing and determine whether
the district court abused its discretion in find- To be sure, plaintiffs did emphasize to the
ing it wanting. We see no abuse of discretion. court that Ascendant was traded on a major
stock exchange, but the court was well within
Plaintiffs argue the district court failed to reason to find this fact alone insufficient to
give due consideration to various factors of show market efficiency and thus predomi-
market efficiency: (1) Ascendant’s listing on nance. After all, the relevant question is
NASDAQ; (2) trading volume; (3) the number whether the market for a particular security is
of market makers and analysts; and (4) stock efficient, because a market can be open and
developed for some securities and not for
others. As the court in Cammer explained,
13
(...continued) “[i]t would be illogical to apply a presumption
insist upon a ‘battle of the experts’ at the class of reliance merely because a security is traded
certification stage,” id., we quoted with approval a within a certain whole market without consid-
statement from the district court’s opinion in this ering the trading characteristics of the individ-
case in defense of considering at least the reliability ual stock itself.” Cammer, 711 F. Supp. at
of expert testimony on market efficiency at the 1281 (internal marks omitted).15
class certification stage. Id. at 323-24 n.6 (citing
Bell, 2004 WL 1490009, at *3-*4). As for trading volume, plaintiffs’ brief in
14 support of class certification did note, in its
Bertulli, of course, recognizes an exception to
this rule, but only for challenges to the power of
statement of facts, Ascendant’s high average
federal courts to entertain the underlying action in
the first instance: “Standing is an inherent pre-
15
requisite to the class certification inquiry; thus, See also id. (“[T]he inquiry in an individual
despite the limited nature of a Rule 23(f) appeal, case remains the development of the market for that
defendants can raise the issue of standing before stock, and not the location where the stock
this court.” Id. trades.”).
8
trading volume. But the relevant indicator is period.17
turnover measured as a percentage of out-
standing shares, see id. at 1286, and plaintiffs Finally, as far as price movement in re-
provided the court with no analysis of this fig- sponse to new, company-specific information
ure. And their effort is no better on appeal, is concerned, plaintiffs addressed this issue in
for they merely repeat the same number with- a sustained way only in the event study pre-
out reference to the total number of outstand- pared by their expert. Yet the district court
ing shares. excluded this evidence as methodologically un-
sound and thus unreliable after concluding
Nor did plaintiffs’ briefing to the district that, once a single day on which no company-
court discuss the presence of market makers specific information was released was exclud-
for Ascendant stock. They contend on appeal, ed from the sample, there was no statistically
however, that the presence of between seven- significant difference between stock-price
teen and twenty-three market makers during movement on so-called information days and
the class period supports a finding of market non-information days. And though plaintiffs
efficiency. Yet, even if we were to consider stress the single-day price decline on the last
this factor on appeal, both the caselaw and day of the class period in response to what
economic literature suggest “the mere number they deem a corrective disclosure, this single
of market makers, without further analysis, has decline on the last day of the class period is
little to do with market efficiency.” Unger, plainly insufficient by itself to show market
401 F.3d at 324.16 efficiency throughout the class period, espe-
cially here where the class period begins as
Instead the relevant information, which early as the day of the IPO itself.18
plaintiffs did not provide, co ncerns “the vol-
ume of shares that they committed to trade,
the volume of shares they actually traded, and 17
We also note that three of the four analysts
the prices at which they did so.” O’Neil v. Ap- who eventually covered the stock and made pur-
pel, 165 F.R.D. 479, 502 (W.D. Mich. 1996). chase recommendations were underwriters of the
Similarly, plaintiffs did not discuss analyst cov- IPO. Cf. William O. Fisher, Does the Efficient
erage in their motion or brief to the district Market Theory Help Us Do Justice in a Time of
court. They do cite on appeal to the expert re- Madness? 54 EMORY L.J. 843, 972 (2005) (sug-
ports submitted in the district court, both of gesting that increasing numbers of sell-side ana-
which discuss analyst coverage, but even if this lysts for stocks, such as Ascendant, that rose and
factor was properly raised in the district court, fell during the Internet bubble should not be re-
it reveals that Ascendant did not have analyst garded as corresponding to an increase in probabil-
coverage for more than a third of the class ity of an efficient market in such stocks, because
analysts were “behaviorally biased” and likely
“contributed to market inefficiency by statistically
biasing price changes”).
16
See also Brad M. Barber et al., The Fraud-
18
on-the-Market Theory and Indicators of Common See Gariety, 368 F.3d at 368 (finding at class
Stock’s Efficiency, 19 J. CORP. L. 285, 307 (1994) certification stage that although drop in price after
(finding that “the number of market makers [does] revelation of company’s insolvency does reflect
not marginally contribute to distinguishing between “the assimilation of market information at its
efficient and inefficient firms”). (continued...)
9
V.
In sum, even if competent evidence could
be marshaled to make a plausible case that As-
cendant common stock traded in an efficient
market such that reliance should be presumed
for the class, this case comes to us with plain-
tiffs’ expert report excluded and their briefing
to the district court devoid of any serious ef-
fort to show market efficiency, so plaintiffs
have not made that case. Accordingly, be-
cause it is their burden to demonstrate that
common issues predominate,19 we find no
abuse of discretion in denying class certifica-
tion. Nor are we persuaded that we should
require that they get a second bite at the class
certification apple; inadequate briefing on an
issue critical to class certification for which a
party bears the burden of proof is no basis for
us to order a repêchage round.
The order denying class certification is
AFFIRMED, and this matter is REMANDED
for further proceedings.
18
(...continued)
grossest level, that single piece of information,
standing alone, does not represent adequate evi-
dence that plaintiffs . . . purchased their shares . . .
in an efficient market”).
19
See O’Sullivan v. Countrywide Home Loans,
Inc., 319 F.3d 732, 737-38 (5th Cir. 2003) (“The
party seeking certification bears the burden of dem-
onstrating that the requirements of rule 23 have
been met.”); Castano, 84 F.3d at 740 (“The party
seeking certification bears the burden of proof.”).
10