DAVID G. CAMPBELL, District Judge.
This is a securities fraud class action brought on behalf of persons who purchased First Solar, Inc. stock between April 30, 2008 and February 28, 2012 (the "Class Period"). First Solar designs and manufactures solar panel modules, and its stock is publicly traded on the NASDAQ Global Market ("NASDAQ"). Plaintiffs have sued First Solar and certain of its officers and directors (collectively, "Defendants"), alleging they made misrepresentations designed to inflate its stock price in violation of Sections 10(b) and 20(a) of the Securities Exchange Act of 1934 (the "1934 Act") and Rule 10b-5.
Plaintiffs move to certify the class. Doc. 156. The motion is fully briefed, and the Court heard oral argument on September 20, 2013. For reasons stated below, the Court will grant the motion.
I. Rule 23 Requirements.
Under Rule 23(a), a district court may certify a class only if (1) it is so numerous that joinder of all members is impracticable, (2) there are questions of law or fact common to the class, (3) the claims of the representative parties are typical of the claims of the class, and (4) the representatives will fairly and adequately protect the interests of the class. Fed. R. Civ. P. 23(a)(1)-(4). The Court must also find that one of the requirements of Rule 23(b) has been met. Plaintiffs rely on Rule 23(b)(3), which requires that questions of law or fact common to the class predominate over any questions affecting only individual class members, and that a class action be superior to other available methods for resolving the controversy. Fed. R. Civ. P. 23(b)(1)-(3).
The Ninth Circuit has not articulated the applicable standard of proof for the Rule 23 requirements, see Conn. Ret. Plans & Trust Funds v. Amgen Inc., 660 F.3d 1170, 1175 (9th Cir. 2011) (Amgen I), aff'd by 133 S.Ct. 1184 (2013) (Amgen II), but at least four circuits have adopted a preponderance of the evidence standard, see Messner v. Northshore Univ. HealthSystem, 669 F.3d 802, 811 (7th Cir. 2012); Novella v. Westchester Cnty., 661 F.3d 128, 148-49 (2d Cir. 2011); Alaska Elec. Pension Fund v. Flowserve Corp., 572 F.3d 221 (5th Cir. 2009); In re Hydrogen Peroxide Antitrust Litig., 552 F.3d 305, 320 (3d Cir. 2008). This standard appears to be the trend in federal courts, Newberg on Class Actions, § 7:21 (Sept. 2013), and "merely requires that [plaintiffs] demonstrate that it is more likely than not that a particular requirement of Rule 23 has been satisfied." Sheperd v. Babcock & Wilcox of Ohio, No. C-3-98-391, 2000 WL 987830, at *1 n.5 (S.D. Ohio Mar. 3, 2000). The Court concludes that Plaintiffs bear the burden of establishing the Rule 23 requirements by a preponderance of the evidence.
Defendants agree that the requirement of Rule 23(a) are satisfied, but argue that Plaintiffs cannot prove that common questions will predominate over individual issues as required by Rule 23(b)(3). Doc. 161 at 5-6. Because the Court must rigorously analyze a class action to ensure it comports with Rule 23, see Wal Mart Stores, Inc. v. Dukes, 131 S.Ct. 2541, 2551 (2011), the Court will first address the Rule 23(a) requirements.
II. Rule 23(a).
Plaintiffs seek certification of the following class:
Doc. 151-6 at 1.
A proposed class satisfies the numerosity requirement if class members are so numerous that joinder would be impractical. Fed. R. Civ. P. 23(a)(1). Plaintiffs provide evidence that "over 1,300 institutional investors alone held First Solar stock during the Class Period," and "over 2.9 billion shares of First Solar stock traded during the Class Period." Doc. 160 at 13 (citing Doc. 158-3 at 5-6). Clearly there are enough class members to make joinder impracticable.
Commonality exists if there are questions of law or fact common to the class. Fed. R. Civ. P. 23(a)(2). "This does not mean merely that they have all suffered a violation of the same provision of the law[.]" Dukes, 131 S. Ct. at 2551. Rather, the common contention underlying the claims "must be of such a nature that it is capable of class-wide resolution — which means that determination of its truth or falsity will resolve an issue which is central to the validity of the claims in one stroke." Id. Because "the complaint alleges a common course of conduct over the entire period directed against all investors, generally relied upon, and violating common statutory provisions, it sufficiently appears that the questions common to all investors will be relatively substantial." Blackie v. Barrack, 524 F.2d 891, 902-03 (9th Cir. 1974) (quoting Harris v. Palm Springs Alpine Estates, Inc., 329 F.2d 909, 914 (9th Cir. 1964)).
A proposed class meets the typicality requirement when "the claims or defenses of the representative parties are typical of the claims and defenses of the class." Fed. R. Civ. P. 23(a)(3). A plaintiff's claim "is typical if it arises from the same event or practice or course of conduct that gives rise to the claims of other class members and his or her claims are based on the same legal theory." Hunt v. Check Recovery Sys., Inc., 241 F.R.D. 505, 511 (N.D. Cal. 2007) (quotation marks and citations omitted). Plaintiffs argue that their claims are typical because they "are based on the same theories of liability, and will be proved by the same evidence" as the class claims. The Court agrees.
D. Adequacy of Representation.
Adequacy requirement is satisfied if the representative parties will fairly and adequately protect the interests of the class. Fed. R. Civ. P. 23(a)(4). This requires that plaintiffs have no conflict of interest with the proposed class and be represented by competent counsel. See Hanlon v. Chrysler Corp., 150 F.3d 1011, 1020 (9th Cir. 1998). There is no evidence that Plaintiffs or their counsel have a conflict of interest with other class members, and the Court previously concluded that Lead Plaintiffs "made a prima facie showing of adequacy," including in their choice of counsel. Doc. 89 at 5. Plaintiffs are adequate representatives of the class.
III. Rule 23(b)(3) Predominance.
"The predominance inquiry of Rule 23(b)(3) asks whether proposed classes are sufficiently cohesive to warrant adjudication by representation." In re Wells Fargo Home Mortg. Overtime Pay Litig., 571 F.3d 953, 957 (9th Cir. 2009) (internal quotation marks and citation omitted). "Considering whether questions of law or fact common to class members predominate begins, of course, with the elements of the underlying cause of action." Erica P. John Fund, Inc. v. Halliburton Co., 131 S.Ct. 2179, 2184 (2011) (internal quotation marks omitted). "The elements of a private securities fraud claim based on violations of § 10(b) and Rule 10b-5 are: `(1) a material misrepresentation or omission by the defendant; (2) scienter; (3) a connection between the misrepresentation or omission and the purchase or sale of a security; (4) reliance upon the misrepresentation or omission; (5) economic loss; and (6) loss causation.'"
"Predominance is a test readily met in certain cases alleging . . . securities fraud," Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 625 (1997), because plaintiffs usually do not need to prove individual reliance and can instead "invoke a rebuttable presumption of reliance based on what is known as the `[fraud on the market]' theory," Halliburton, 131 S. Ct. at 2185; see also Basic, Inc. v. Levinson, 485 U.S. 224, 241-49 (1988) (endorsing fraud on the market theory). That theory relies on "the principle that the market price of a security traded in an efficient market reflects all public information and therefore that a buyer of the security is presumed to have relied on the truthfulness of the information in purchasing the security." Amgen I, 660 F.3d at 1172; see also Basic, 485 U.S. at 241-42. It "facilitates class certification by recognizing a rebuttable presumption of classwide reliance on public, material misrepresentations when shares are traded in an efficient market." Amgen II, 133 S. Ct. at 1193. "Absent the [fraud on the market] theory, the requirement that Rule 10b-5 plaintiffs establish reliance would ordinarily preclude certification of a class action [under Rule 23(b)(3)] because individual reliance issues would overwhelm questions common to the class." Amgen II, 133 S. Ct. at 1193 (citing Basic, 485 U.S. at 242).
A. Plaintiffs' Exclusive Reliance on Fraud on the Market.
Plaintiffs argue that the predominance requirement is satisfied by the simple fact that they will never seek to prove individual class member reliance on false statements of Defendants. Instead, they will show reliance through the fraud on the market theory. Because fraud on the market (and therefore reliance) will either be proved or fail for the class as a whole, common issues necessarily will predominate. As a result, Plaintiffs assert, the Court need not make any inquiry into the efficiency of the market for First Solar stock at the class certification stage — the Court can conclude with confidence that common issues will predominate simply because Plaintiffs rely exclusively on the fraud on the market theory of reliance. This argument has some logical force, but the Court cannot accept it.
First, the argument is contrary to established Rule 23 case law. Every securities fraud class action based on the fraud on the market theory disclaims any intent to prove reliance on the part of individual class members. Such cases are premised on the fact that the class representatives will prove reliance solely through fraud on the market. Courts nonetheless require class plaintiffs to prove market efficiency at the class certification stage. See Dukes, 131 S. Ct. at 2552 n.6 ("To invoke [the fraud on the market] presumption, the plaintiffs seeking 23(b)(3) certification must prove that their shares were traded on an efficient market, an issue they will surely have to prove again at trial in order to make out their case on the merits." (emphasis in original; citation omitted)); Halliburton, 131 S. Ct. at 2185 (at class certification stage, "plaintiffs must demonstrate. . . that the stock traded in an efficient market"); Amgen I, 660 F.3d at 1175 ("the district court was correct to require [plaintiff] to prove at the class certification stage that the market for [defendants'] stock was efficient"); In re Diamond Foods, Inc., Sec. Litig., No. C 11-05386 WHA, 2013 WL 1891382, at *10 (N.D. Cal. May 6, 2013) ("a showing of market efficiency is clearly required at the class certification stage in an action relying on the [fraud on the market] theory").
Second, Rule 23 requires more than careful pleading. Under Plaintiffs' theory, the predominance question as to reliance would begin and end at the class complaint. If the complaint relied exclusively on the fraud on the market theory, no further factual inquiry would be needed. The Supreme Court has made clear, however, that "Rule 23 does not set forth a mere pleading standard." Dukes, 131 S. Ct. at 2551 (citation omitted). "A party seeking class certification must affirmatively demonstrate his compliance with [Rule 23.]" Id. Thus, courts at the class certification stage engage in "rigorous analysis" that entails "some overlap with the merits of the plaintiff's underlying claim." Id. Plaintiffs' pleadings-only approach would conflict with these clear Supreme Court instructions.
Third, Plaintiffs cite no case where common issues were found to predominate on the issue of reliance solely because the plaintiffs relied exclusively on a fraud on the market theory. Plaintiffs ask this Court to break new ground. The Court declines the invitation.
Plaintiffs cite Halliburton, Amgen I, and Amgen II in support of their argument. Halliburton held that plaintiffs need not establish loss causation in order to invoke the fraud on the market theory of reliance at the certification stage. 131 S. Ct. at 2183-85. In Amgen I, the Ninth Circuit held that plaintiffs need not prove materiality to invoke the fraud on the market theory, 660 F.3d at 1175-76, and the Supreme Court affirmed in Amgen II, 133 S.Ct. 1184. These cases do not hold that plaintiffs can satisfy Rule 23(b)(3) by pleading exclusive reliance on the fraud on the market theory.
In sum, the Court is not persuaded by Plaintiffs' novel argument. Plaintiffs must prove at this stage of the litigation that questions of law or fact common to the class predominate over questions affecting only individual members.
B. Plaintiffs Rule 23(b)(3) Proof.
Because reliance is an issue on which individual issues can predominate in a securities fraud case, and the fraud on the market theory is designed to prove reliance on a class-wide basis, Plaintiffs must prove at the class certification stage that they can utilize the fraud on the market theory. To do so, plaintiffs must show that the alleged misrepresentations were publicly known, that the stock transactions occurred between the time the misrepresentations were made and the time the truth was revealed, and that the stock traded on an efficient market. Halliburton Co., 131 S. Ct. at 2185. Defendants' sole argument in opposing class certification is that Plaintiffs have failed to show that First Solar stock traded on an efficient market. To prove market efficiency, Plaintiffs submit that the NASDAQ is entitled to a presumption of efficiency, and that the five factors identified in Cammer v. Bloom, 711 F.Supp. 1264, 1285-87 (D.N.J. 1989), also show market efficiency.
1. Presumption of Market Efficiency.
NASDAQ is "one of the two largest stock exchanges in the United States, the largest electronic-equity securities trading market in the United States, and one of the largest stock exchanges in the world." Lumen v. Anderson, 280 F.R.D. 451, 459 (W.D. Mo. 2012). Because generally "the market price of shares traded on well-developed markets reflect all publicly available information," Basic, 485 U.S. at 246, some courts have found that trading on a major exchange weighs in favor of finding that the stock traded in an efficient market. See, e.g., In re DVI, Inc. Sec. Litig., 639 F.3d 623, 634, (3d Cir. 2011) ("the listing of a security on a major exchange such as the NYSE or NASDAQ weighs in favor of a finding of market efficiency"), abrogated on other grounds by Amgen II, 133 S.Ct. 1184; Lumen, 280 F.R.D. at 459 ("It would be remarkable for a court to conclude NASDAQ is not an efficient market — which is why securities traded on NASDAQ are often presumed to be traded on an efficient market." (internal quotation marks, brackets, and citations omitted)); In re Juniper Networks, Inc. Sec. Litig., 264 F.R.D. 584, 591 (N.D. Cal. 2009) ("Plaintiffs made a prima facie showing that the [fraud on the market] presumption of reliance applied because . . . stock was actively traded on an efficient market — the NASDAQ."). The Supreme Court's decision in Basic appears to support such a presumption. See 485 U.S. at 246 ("In drafting [the 1934] Act, Congress expressly relied on the premise that securities markets are affected by information, and enacted legislation to facilitate an investor's reliance on the integrity of those markets. . . . Recent empirical studies have tended to confirm Congress' premise that the market price of shares traded on well-developed markets reflects all publicly available information, and, hence, any material misrepresentations." (citations omitted)).
Defendants argue that there is no presumption of market efficiency for securities listed on a major national exchange. In support, Defendants cite George v. China Auto. Sys., Inc., No. 11 Civ. 7533 (KBF), 2013 WL 3357170 (S.D.N.Y. July 3, 2013); Dean v. China Agritech, No. CV 11-01331-RGK (PJWx), 2012 WL 1835708, at *6-8 (C.D. Cal. May 3, 2012); In re PolyMedica Corp. Sec. Litig., 453 F.Supp.2d 260, 278-79 (D. Mass. 2006) (PolyMedica II); In re Northfield Labs., Inc. Sec. Litig., 267 F.R.D. 536 (N.D. Ill. 2010); and Bell v. Ascendant Solutions, Inc., 422 F.3d 307, 316 (5th Cir. 2005). Of these cases, however, only PolyMedica II considered the significance of a stock's listing on a national exchange, and even there the district court noted "that listing on such an exchange undisputably improves the market structure for trading in a particular stock" and "that one would be hard-pressed to deny the relevance of this fact in an efficiency analysis." 453 F. Supp. 2d at 266 (citing O'Neil v. Appel, 165 F.R.D. 479, 504 (W.D. Mich. 1996)). Bell noted only that the district court "was well within its reason to find this fact alone insufficient to show market efficiency," 422 F.3d at 315, and the other cited cases appear to have ignored the fact that the securities were listed on large national exchanges, see, e.g., China Auto. Sys., 2013 WL 3357170; China Agritech, 2012 WL 1835708; Northfield Labs., 267 F.R.D. 536.
At oral argument, Defendants represented that Binder v. Gillespie, 184 F.3d 1059 (9th Cir. 1999), rejected any presumption of market efficiency based on the size of the exchange. In Binder, the stock initially traded exclusively on the over-the-counter market and then later on the Boston Stock Exchange ("BSE"). Id. at 1064-65. The district court determined that plaintiff could not invoke the fraud on the market presumption with respect to the period in which the stock traded exclusively on the over-the-counter market, and thus decertified the class for this period. Id. For the period in which the stock traded on the BSE, the district court noted that there was no showing that the BSE was an inefficient market, but it "nonetheless decertified the class for the remaining period . . . on the ground that [plaintiff] could not prove causation." Id. at 1065. The Ninth Circuit affirmed the district court, but did not discuss the appropriateness of employing a presumption of market efficiency to established markets. Binder therefore does not support Defendants' position.
In keeping with Basic and the other cases cited in the first paragraph of this section, the Court concludes that the trading of First Solar stock on NASDAQ — a major, well-developed stock exchange — weighs in favor of finding market efficiency. Defendants have not "identified any authority, binding or otherwise, that has held that common shares traded on the NASDAQ are not traded in an efficient market." Diamond Foods, 2013 WL 1891382 at *10. Moreover, Defendants' expert, Dr. Gompers, agrees that "[m]ost of the time, . . . stocks traded on large national exchanges are likely to be efficient." Doc. 165-1 at 15.
2. The Five
The district court's decision in Cammer outlined a five-factor test for market efficiency that has been widely utilized. The factors are "first, whether the stock trades at a high weekly volume; second, whether securities analysts follow and report on the stock; third, whether the stock has market makers and arbitrageurs; fourth, whether the company is eligible to file SEC registration form S-3, as opposed to form S-1 or S-2; and fifth, whether there are `empirical facts showing a cause and effect relationship between unexpected corporate events or financial releases and an immediate response in the stock price.'" Binder, 184 F.3d at 1065 (quoting Cammer, 711 F. Supp. at 1286-87); see also Miller v. Thane Int'l, Inc., 615 F.3d 1095, 1103 (9th Cir. 2010) (noting that the Cammer test "was developed in support of a judicial presumption allowing plaintiffs to avoid a significant obstacle in certifying a class in securities fraud litigation."). It has been suggested that "[t]he first four Cammer factors indirectly assess market efficiency by evaluating whether the attributes of the market for that security are conducive to informational efficiency." In re Countrywide Fin. Corp. Sec. Litig., 273 F.R.D. 586, 613 (C.D. Cal. 2009) (citing PolyMedica I, 432 F.3d at 18; In re Xcelera.com Sec. Litig., 430 F.3d 503, 511 (1st Cir. 2005)). Defendants submit that the fifth factor is the most important because it alone "explores whether an important result of an efficient market exists," as opposed to simply "relying on circumstantial evidence that a security's market is conducive to efficiency[.]" Countrywide, 273 F.R.D. at 614 (emphasis in original).
a. Four of the
Plaintiffs provide evidence that the first, second and fourth Cammer factors, and at least part of the third, are satisfied: First Solar stock traded at a high weekly volume, a significant number of securities analysts followed and reported on the stock, there were at least 31 market makers for the stock as well as some 1,200 institutional investors who traded on public information, and First Solar was eligible to file an S-3 registration statement throughout the Class Period and did so on April 2, 2009. Doc. 160 at 18-19 (citing Doc. 158-3 at 6-23). In addressing the fifth factor — whether a cause and effect relationship existed between unexpected corporate events and a quick response in First Solar's stock price — Plaintiffs point to five days within the Class Period where the price of First Solar stock responded to a company-specific news event (Doc. 160 at 20 (citing Doc. 158-3 at 24-26)), evidence that statistically significant price movements occurred more often on days with news than without news (Doc. 160 at 20 n.9 (citing Doc. 158, ¶¶ 3-5, and Doc. 158-2 at 59-233)), and evidence that statistically significant price movements followed 13 of 16 First Solar earnings announcements issued during the Class Period (Doc. 160 at 20 n.9 (citing Doc. 158, ¶¶ 3-4).
Defendants do not dispute Plaintiffs' evidence on the first, second, or fourth Cammer factors. Defendants focus exclusively on the third and fifth factors, arguing that Plaintiffs' third-factor showing is "incomplete" and that their fifth-factor showing is unsupported. Id.
b. The Third
"The existence of market makers and arbitrageurs would further provide a mechanism through which the market could be expected to receive information and fully incorporate it into the stock price of a security, as these individuals `would react swiftly to company news and reported financial results by buying or selling stock and driving it to a changed price level.'" Diamond Foods, 2013 WL 1891382, at *8 (quoting Cammer, 711 F. Supp. at 1286-87); see also Countrywide, 273 F.R.D. at 612 ("the presence of `arbitrageurs' — those who quickly exploit (usually small) price differences among different markets — help bring prices across markets quickly into line with one another, enabling different markets to reflect the same information."). As noted, Plaintiffs show that there were 31 market makers for First Solar stock during the Class Period as well as more than 1,200 investors. Defendants contend that Plaintiffs ignore arbitrageurs.
Defendants assert that arbitrageurs were unable to participate in the market effectively during the Class Period. Relying on the opinion of Dr. Gompers, Defendants submit that because short-selling was expensive and difficult during part of the Class Period (id. at 18 (citing Doc. 162-2, ¶¶ 54, 57, 59)), "the market for short-selling the stock became constrained, and the stock price could not fully and rapidly reflect all negative information during that portion of the class period" (id. at 19 (citing Doc. 162-2, ¶ 54)).
Defendants submit that the arbitrage market for First Solar stock became constrained in response to an "extraordinarily high" interest in short selling that in turn caused the cost of borrowing First Solar stock for short-selling purposes to "skyrocket." Doc. 161 at 18. Dr. Gompers identified an eleven-week period, from June 6 to August 25, 2011, during which there was both significant interest in short-selling and an increase in borrowing fees. Id. (citing Doc. 162-2, ¶¶ 54, 57, 59). Dr. Gompers also found that the amount of time it took "to cover (i.e., repurchase) all shares sold short" increased from one day at the beginning of the Class Period to 18.9 days as of July 29, 2011. Doc. 162-2, ¶ 54. Due to significant short-selling interest, high loan fees, and the additional days required to cover short positions, Dr. Gompers opines that "arbitrage in First Solar stock was constrained over sizeable stretches of the putative class period." Id., ¶ 60.
To show the relevancy of this short-sale analysis, Defendants rely primarily on PolyMedica II, 453 F. Supp. 2d at 273, a case in which the court noted that impediments to short selling constitute "indirect evidence" of market inefficiency. Doc. 161 at 13, 19. The defendant in PolyMedica II offered evidence that short-selling was difficult during the contested period because (1) the interest in short-selling began to "skyrocket," (2) "finding shares to short became very difficult," and (3) "the transaction costs for short selling . . . became extraordinarily high." Id. at 273-74. The court also noted, however, that "[t]he tangible effect of these constraints on short selling is seen in yet another piece of evidence proffered by [defendant]: violations of put-call parity." Id. at 274. This evidence was significant because "[u]sing derivative securities called `puts' and `calls' (together, `options') in combination with short sales, arbitrageurs can make guaranteed profits if certain conditions are present," that is, if put-call parity is violated. Id. at 274-75.
The Court has three concerns about Defendants' short-sale argument. First, it is unclear from PolyMedica II whether impediments to short selling demonstrate market inefficiency absent a corresponding put-call disparity, and Defendants present no evidence of such disparity in this case. Second, while arguing that there were serious impediments to short selling First Solar stock, Defendants also note admit that "of all the stocks traded on the NASDAQ, First Solar was one of the most heavily `shorted' stocks during the proposed class period." Doc. 161 at 12 (citing Doc. 162-2, ¶¶ 53-60). Third, Defendants' evidence suggests that short selling was constrained for only 11 weeks out of the four-year Class Period.
Despite these misgivings, Defendants' short-sale evidence warrants consideration in the class certification analysis. It constitutes some indirect evidence of market inefficiency and thus should be weighed against the indirect evidence of market efficiency Plaintiffs present.
c. The Fifth
It is "helpful to a plaintiff seeking to allege an efficient market to allege 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. This "is the essence of an efficient market and the foundation for the fraud on the market theory." Id.
Plaintiffs point to five days within the Class Period in which a statistically significant price reaction at the 1% level
i. Five Event Days.
According to Defendants, Dr. Steinholt's study of only five event days is insufficient proof of market efficiency because those five days represent a small portion of the Class Period. Id. at 14-15. They point to China Automotive Sys., 2013 WL 3357170 at *12, which found that a "showing that only seven out of sixteen days resulted in a market reaction is an insufficient foundation upon which to pronounce market efficiency." It does not appear, however, that the China Automotive finding was based on the 22-month class period, as Defendants suggest (Doc. 161 at 14), but instead on the less than 50% showing of statistically significant stock price reactions in the right direction on predefined dates expected to impact stock pricing. Id. at *11-12. Defendants also point to a footnote in Countrywide, 273 F.R.D. at 618 n.96, describing the problems of subjective determinations in event studies and noting that ambiguity in such studies "can be reduced by increasing the sample size (number of `observations'), ensuring the sample is representative (e.g., by studying a random sample of days with both large and small returns), and objectively defining `new' information." Additionally, Defendants note that the five event days are not evenly distributed throughout the Class Period. Doc. 161 at 14. Defendants point to PolyMedica II's conclusion that plaintiff's expert's selection of "the largest stock price drops" — five days out of about 160 trading days — "proves nothing." 453 F. Supp. 2d at 169-70. But Dr. Steinholt appears to have done better than the expert in PolyMedica II because his event study considered statistically significant price movements (Doc. 158-2, ¶¶ 28-33), whereas the PolyMedica II expert examined general price fluctuations only, 453 F. Supp. 2d at 169-70.
Defendants criticize Dr. Steinholt's event selection "because he did not use articulable objective criteria to select news events[.]" Doc. 161 at 17-18 (citing Countrywide, 273 F.R.D. at 618 ("the events for study should be selected using criteria that are as objective as possible."). Defendants presume that Dr. Steinholt selected some of the events, and the CEO resignation announcement in particular, "only because it coincided with a statistically significant stock reaction" (id. at 18). Courts have acknowledged that "unless the expert uses articulable objective criteria [to select events], it is difficult to evaluate the probative value of expert evidence without evaluating also the expert's own credibility." Countrywide, 273 F.R.D. at 618. Additionally, Defendants argue that Dr. Steinholt's event selection is insufficient because he did not consider any of the alleged misstatements at issue in this case. Doc. 161 at 15.
These criticisms have some merit. They suggest that Dr. Steinholt's event days were not selected with fully objective criteria, and that five days during a four-year class period certainly is not comprehensive. But Defendants do not dispute that statistically significant price reactions at the 1% confidence level followed five significant announcements regarding First Solar's financial condition. Thus, although Defendants have provided reason to believe that Plaintiffs' event days are not fully objective and are somewhat incomplete, they have not shown the data to be inaccurate.
As to Plaintiffs' comparison of "news" versus "non-news" days, Defendants argue that Plaintiffs' counsel, who is not an expert in market efficiency, made the comparison and Plaintiffs cannot supplement their expert report after the close of discovery. Doc. 161 at 16. The Court disagrees with both arguments. Instead of offering an expert opinion, Plaintiffs' counsel performed calculations on the data in Dr. Steinholt's report. See Doc. 164 at 12-13 n.7. The Court could have performed the same calculations given the data in the report, and does not consider such calculations to constitute new expert opinions.
With respect to Plaintiffs' assertion that First Solar stock made significant price moves in response to 81.25% of the earnings reports released during the Class Period, Defendants argue that Plaintiffs must show that the stock traded efficiently throughout the Class Period. Doc. 161 at 16. Although the Court generally agrees, evidence of market efficiency on some days can be indicative of market efficiency on other days, and many days are simply flat as to price and news, leaving little with which to prove market efficiency. The Court does not agree with Defendants' suggestion that the fifth Cammer factor requires a day-by-day showing of market efficiency for class certification. Cammer states that "it would be helpful" for a plaintiff to "allege empirical facts" as to the fifth factor, but Cammer does not require empirical facts as to every trading day. 711 F. Supp. at 1287.
Additionally, Defendants contend that Lumen does not support Plaintiffs' assertion that 81.25% "is good enough" because Plaintiffs have not shown statistically significant movement on every "news" day as was shown in Lumen. In considering the fifth Cammer factor, Lumen noted that the defendants submitted evidence that on 35% of the days in which the stock exhibited statistically significant stock movement, there was no company-specific news. 480 F.R.D. at 461. Lumen found that this also "means that 65% of the time statistically significant stock price moves are related to company-specific news." Id. Lumen did not find that there was statistically significant movement on every "news" day. The Court also notes that in DVI, 639 F.3d at 635, the Third Circuit concluded that the district court did not err in finding that 60% and 65% correlations between news releases and stock price changes weighed in favor of market efficiency. Accordingly, the Court cannot accept Defendants' assertion that because 81.25% is not 100%, Plaintiffs have made an insufficient showing with respect to Cammer factor five.
Cammer Factor Five Evidence.
Defendants point to two instances in which the price of First Solar stock did not have a statistically significant response to unexpected good news: (1) First Solar issued an unexpectedly positive quarterly earnings release on April 30, 2008, and the residual return following the release was .76%, and (2) First Solar issued an unexpectedly positive quarterly earnings release on July 30, 2008, and the residual return following the release was 1.06%. Doc. 161 at 15 (citing Doc. 162-2, ¶¶ 28, 40). Dr. Gompers found that the stock price reactions to these two earnings announcements did not result in statistically significant reactions (Doc. 162-2, ¶ 40), and that the reactions "were at odds with what you would expect from a full and rapid incorporation of new-value relevant information" (Doc. 162-5 at 18:16-18). This evidence, although worthy of consideration, is less than conclusive. See Lumen, 280 F.R.D. at 461 (noting that Dr. Gompers' opinion that the "stock did not incorporate public information . . . [is] wanting" because it "is based on only two days on which [the stock] experienced no significant change in stock price despite the announcement of important news. From this, he concludes the market was inefficient. However, there is no comparison to the number of times there were significant price changes following important news.").
Defendants argue that Dr. Steinholt's "event study shows two back-to-back days of statistically significant stock reactions on December 16 and 17, 2008" when "there was no value relevant news on either day." Doc. 161 at 20. Defendants submit that "[w]hen the market for a stock is operating efficiently, that should not happen" (id.), and that "the likelihood of it happening by pure chance is less than one one-hundredth of one percent (0.01%)" (id.).
Defendants also argue that evidence of serial correlation demonstrates market inefficiency. Defendants submit that "[i]n an efficient market, an investor cannot predict future stock prices based on past movements." Doc. 161 at 20 (citing PolyMedica II, 453 F. Supp. 2d at 276-77; Countrywide, 273 F.R.D. at 614). Defendants then point to their expert's finding of serial correlation in First Solar stock prices from June 6 to August 25, 2011. Id. (citing Doc. 162-2, ¶ 63). Serial correlation may be evidence of market inefficiency because if past stock price returns tend to predict current or future returns, there is a delay in the adjustment of stock prices to new information and thus "the direction in which price moves today is a statistically significant predictor of the direction in which it will move tomorrow[.]" PolyMedica II, 453 F. Supp. 2d at 277. Although in PolyMedica II the court found that evidence of serial correlation, in combination with barriers to short selling, rebutted plaintiff's weak showing of market efficiency, 453 F. Supp. 2d at 278, the First Circuit in PolyMedica I "ma[de] no judgment on the relevance of [the serial correlation test], or the appropriateness of the court's consideration of this evidence during the class-certification proceedings," 432 F.3d at 18 n.21. Defendants have made a showing of serial correlation for only 11 weeks out of an almost four-year Class Period, and some courts have found that "the presence of serial correlation is not itself determinative of inefficiency." Countrywide, 273 F.R.D. at 615.
iv. Factor Five Conclusion.
Plaintiffs have presented evidence of efficient-market responses on five specific days and evidence of meaningful stock price responses on 81.25% of the days when earnings releases occurred. Defendants criticize this evidence as nonobjective and incomplete, but their own empirical evidence is limited to two days when First Solar stock failed to respond significantly to unexpected good news, two days when it moved significantly in response to no news, and 11 weeks of serial correlation, a factor discounted by at least one court. The Court views this evidence as essentially a stand-off on the fifth Cammer factor. If forced to declare a victor, the Court likely would choose Plaintiffs because of the 81.25% correlation between First Solar earnings reports and significant price moves, but the empirical battle is too close to call on the basis of such limited evidence. The Court has no difficulty, however, reaching a conclusion on the broader question of market efficiency.
D. Market Efficiency Conclusion.
Plaintiffs have shown by a preponderance of the evidence that First Solar stock traded on an efficient market. That evidence includes the following undisputed facts: First Solar stock traded on the NASDAQ, a major and well-developed securities market; the stock traded at high weekly volumes; securities analysts followed and reported on the stock; the stock had 31 market makers during the Class Period and more than 1,200 institutional investors; and First Solar was eligible to file and did file SEC registration form S-3. In addition, the First Solar stock price made statistically significant moves on five specific days when noteworthy First Solar news was released and on 81.25% of the days when earnings were released. Defendants have presented evidence that arbitrageurs were likely constrained during a portion of the Class Period, that First Solar stock failed to respond significantly to unexpected good news on two days, that it moved significantly in response to no news on two other days, and that serial correlation is apparent in 11 weeks of the Class Period.
In weighing this evidence, the Court finds it significant that Plaintiffs have shown without dispute that three of the five Cammer factors are satisfied, that another is partially satisfied, and that the stock traded on the NASDAQ. Defendants' evidence that the third Cammer factor is partially unsatisfied, and the stand-off on the fifth Cammer factor, simply do not outweigh Plaintiffs' evidence, even if the fifth factor is considered most important. The Court therefore concludes that Plaintiffs have shown market efficiency by a preponderance of the evidence, that the fraud on the market theory is warranted in this case, and that common issues will predominate in light of the fraud on the market theory of reliance.
IV. Rule 23(b)(3) Superiority.
The superiority inquiry asks whether "a class action is superior to other available methods for fairly and efficiently adjudicating the controversy." Fed. R. Civ. P. 23(b)(3). Rule 23(b)(3) identifies four factors a court must consider in determining superiority, see Fed. R. Civ. P. 23(b)(3)(A)-(D), "so that cases allowed under subdivision (b)(3) are those that can be adjudicated most profitably on a representative basis[.]" Zinser v. Accufix Research Inst., Inc., 253 F.3d 1180, 1190 (9th Cir. 2001). Plaintiffs submit that "there are no other securities actions pending against defendants; litigating all claims in this District (where First Solar is headquartered) is far more efficient than in the home forum of each class member; and because class-actions are the rule, rather than the exception, for cases under the PSLRA, there will be few, if any, difficulties in managing this case as a class action." Doc. 160 at 23. Defendants do not contest Plaintiffs' suggestion that a class action is superior.