Learn About the Law
Get help with your legal needs
FindLaw’s Learn About the Law features thousands of informational articles to help you understand your options. And if you’re ready to hire an attorney, find one in your area who can help.
MONROE COUNTY EMPLOYEES' RETIREMENT SYSTEM and Roofers Local No. 149 Pension Fund, Individually and on Behalf of All Others Similarly Situated, Plaintiffs, v. The SOUTHERN COMPANY, Thomas A. Fanning, Art P. Beattie, Edward Day, VI, G. Edison Holland, Jr., John C. Huggins, and Thomas O. Anderson, Defendants.
OPINION AND ORDER GRANTING MOTION FOR CLASS CERTIFICATION AND APPOINTMENT OF CLASS REPRESENTATIVES AND CLASS COUNSEL
This is a federal securities fraud class action brought against defendant The Southern Company (the “Southern Company” or the “Company”) and Individual Defendants Thomas A. Fanning, Art P. Beattie, Edward Day, VI, G. Edison Holland, Jr., John C. Huggins, and Thomas O. Anderson (collectively the “Defendants”). Roofers Local No. 149 Pension Fund (the “Roofers Local No. 149”) and Monroe County Employees' Retirement System (the “Monroe County” and, together with Roofers Local No. 149, the “Plaintiffs”), as putative class representatives, seek damages for Defendants' alleged violations of Section 10(b) of the Securities Exchange Act of 1934 (the “Exchange Act”), 15 U.S.C. § 78j(b), Rule 10b-5 promulgated thereunder, and Section 20(a) of the Exchange Act. The operative complaint in this action is the Consolidated Complaint for Violations of the Federal Securities Laws, filed on June 12, 2017. [Doc. 28] (the “Complaint”).
This matter is before the Court on Plaintiffs' Motion for Class Certification. [Doc. 77] (the “Plaintiffs' Class Certification Motion”). The motion has been fully briefed, and with the benefit of oral argument on May 21-22, 2019, the Court enters the following Order.
I. FACTUAL BACKGROUND
A detailed description of the allegations in this case is set forth in the March 29, 2018, Order granting in part and denying in part Defendants' motion to dismiss the Complaint. [Doc. 43] (“MTD Order”). To summarize, Plaintiffs allege the following:
In January 2009, Mississippi Power Company (“Mississippi Power”), a wholly-owned subsidiary of Southern Company, announced that it was planning to construct a “clean coal” power plant in Kemper County, Mississippi (the “Kemper Plant”). ¶33.1 During 2009, the U.S. Internal Revenue Service (“IRS”) certified the allocation of $133 million in tax credits to Mississippi Power for the construction of the Kemper Plant so long as Defendants completed the plant by its proposed commercial operation date in May 2014 (the “May 2014 COD”). ¶¶4, 35. In June 2010, the Mississippi Public Service Commission (“PSC”), which regulates Mississippi public utilities, authorized the acquisition, construction, and operation of the Kemper Plant and approved a ratepayer-funded allowance to finance the $2.88 billion construction costs incurred through the May 2014 COD. ¶¶32, 36. The PSC's authorization provided that any costs incurred beyond $2.88 billion or subsequent to the May 2014 COD could not be passed on to Mississippi Power's ratepayers. ¶36. In addition, the South Mississippi Electric Power Association made a $150 million deposit toward a 15% ownership interest in the Kemper Plant, which Southern Company was required to repay with interest if the plant was not completed by the May 2014 COD. ¶¶12, 172. Lastly, Southern Company had a contract with Treetop Midstream Services, LLC (“Treetop”), whereby Treetop would purchase the CO2 by-product generated by the Kemper Plant and construct a pipeline to offload the CO2 by-product, again provided that the plant was timely constructed. ¶12. As such, it was imperative that the Kemper Plant be completed on time. ¶5.
Plaintiffs allege that throughout the construction process, Defendants repeatedly assured the public and investors that construction of the Kemper Plant was “ ‘on target,’ ” “ ‘on schedule,’ ” “ ‘70 percent complete,’ ” “ ‘75 percent complete,’ ” “exceedingly well-built and well organized,” reaching specific component milestones, and would be completed by the May 2014 COD.2 Monroe Cty. Emps.' Ret. Sys. v. S. Co., No. 1:17-CV-241-MHC, 2018 WL 1558577, at *7-*23 (N.D. Ga. Mar. 29, 2018); see also ¶¶116, 118-119, 122-123, 125-127, 130, 132-137, 140-141, 143, 145, 149-151, 153, 155, 158 (detailing the remaining alleged false statements).
Plaintiffs allege that Defendants knew or recklessly disregarded that their Class Period statements were misleading or omitted information necessary to make their statements not misleading. Plaintiffs allege that from the outset and continuing through the end of the Class Period, Defendants knew or recklessly disregarded that the Kemper Plant was not “on schedule” or “on target,” was not “70 percent” or “75 percent complete” when they so stated; that construction was not “well organized”; that specific component milestones had not been achieved; and that the “[t]he May 2014 COD was impossible to achieve.” See, e.g., ¶124(e) (alleging Defendants' May 15, 2012 statement that “we'll have the first heat to the gasifier in October of next year” (¶116) was false and misleading in part because Defendants knew or recklessly disregarded that “[t]he May 2014 COD was impossible to achieve due to major delays in the installation of the gasifier”); Monroe Cty. Emps.' Ret. Sys. v. S. Co., 333 F. Supp. 3d 1315, 1324 (N.D. Ga. 2018) (finding Plaintiffs' allegations of scienter sufficient, as “ ‘[Brett] Wingo [a former employee and whistleblower] said that as soon as he learned in February 2012 that there were refractory failures on the gasifier, he knew that there was “no way in hell” that Southern Company could meet the May 2014 deadline’ ”) (emphasis in original) (quoting ¶61); see also ¶¶115(e), 131(f)-(h), 144(f)-(h), 159(f)-(h) (alleging all of Defendants' Class Period misstatements were misleading in part because Defendants knew or recklessly disregarded that “[t]he May 2014 COD was impossible to achieve”).
Considering Defendants' motion to dismiss, Judge Cohen held that “Plaintiffs' Complaint is replete with allegations of facts which, if found to be true, demonstrate that the statements were false or misleading.” Monroe Cty., 2018 WL 1558577, at *22. For example, while Defendants misleadingly stated “that the gasifier was installed as of September 13, 2012,” “the first gasifier was not delivered until January of 2013.” Id. Plaintiffs also presented internal Company emails to the Court at the May 22, 2019, hearing that Plaintiffs contend support the Complaint's allegations. For example, an April 2012 email to defendant Anderson discussed a “revised completion date” for the Kemper Plant and noted that, as of April 2012, the Kemper Plant was “not on time and [was] over budget.” [Doc. 128-14] (“Plaintiffs' Ex. 11”) at 4. The email to defendant Anderson also discussed what steps, if any, were being “tak[en] to bring the project back on schedule.” Id. Further, in June 2012, Wingo informed his superior that a change in materials “[would] result in a 40 week delay” and emphasized that “there [was] no getting around it.” Id. at 5. Plaintiffs allege that Defendants continued to make false and misleading statements regarding the construction of the Kemper Plant and actively sought to conceal the true state of the construction. See, e.g., id. at 7 (August 8, 2012, email from defendant Day instructing his employees that nothing regarding the Kemper Plant schedule should be shared publicly and that any schedule or budget information should be approved by Defendants Day, Huggins, and Anderson before public dissemination).
Finally, Plaintiffs allege that when Defendants could no longer conceal the schedule and construction issues at the Kemper Plant, they disclosed what they had known from the very start of the Class Period – the Kemper Plant was not “on schedule” or “on track” and the May 2014 COD was impossible to achieve. ¶45. In fact, as of the filing of the Complaint in June 2017, “Defendants announced that the Kemper Plant was not yet fully commercially operable and would cost nearly three times its original construction cost limit.” Monroe Cty., 2018 WL 1558577, at *1. Since then, Defendants have announced that the Kemper Plant will never operate as a “clean coal” facility. [Doc. 128-14 at 10]. In addition to the other government and private investigations into the Kemper Plaint, in April 2019, Defendants disclosed that the Department of Justice had opened an investigation concerning the Kemper Plant. Id. at 11.
II. PROCEDURAL BACKGROUND
On April 11, 2017, the Court appointed Roofers Local No. 149 as Lead Plaintiff and Robbins Geller Rudman & Dowd LLP (“Robbins Geller”) as Lead Counsel. [Doc. 22]. On June 12, 2017, Plaintiffs filed their Complaint alleging that between April 25, 2012, and October 30, 2013, Defendants violated Sections 10(b) and 20(a) of the Exchange Act by issuing false and misleading statements about the construction of the Kemper Plant.
On March 29, 2018, the Court granted in part and denied in part Defendants' motion to dismiss [Doc. 43] (“MTD Order”). On April 6, 2018, the Court granted Defendants' motion for clarification of the MTD Order to clarify that Defendants' motion to dismiss Plaintiffs' Section 20(a) claim was denied as to all Defendants and that all of the Individual Defendants remained in the case. [Doc. 47]. On April 26, 2018, Defendants moved for reconsideration of the MTD Order. [Doc. 51]. On May 23, 2018, Defendants moved for certification of the MTD Order, pursuant to 28 U.S.C. § 1292(b), and requested a stay pending appeal. [Doc. 57]. On August 10, 2018, the Court denied Defendants' motion for reconsideration and Defendants' motion for interlocutory appeal and stay. [Doc. 68].
On September 24, 2018, Plaintiffs moved for class certification and for appointment of class representatives and class counsel. [Doc. 77]. In support of their motion, Plaintiffs submitted the expert opinion of Professor Steven P. Feinstein, Ph.D., CFA (“Professor Feinstein”), on market efficiency and damages. On February 4, 2019 [Doc. 77-2], Defendants opposed class certification and moved to exclude certain of Professor Feinstein's opinions under Federal Rule of Evidence 702. [Docs. 106, 109]. In support of their submissions, Defendants submitted the opinions of Professor Paul A. Gompers, Ph.D. (“Professor Gompers”) [Doc. 106-2]. On March 29, 2019, Plaintiffs moved to exclude the opinions of Professor Gompers [Doc. 114].
On May 21-22, 2019, the Court conducted an evidentiary hearing on Plaintiffs' motion for class certification and the parties' competing motions to exclude. During the hearing, the Court heard testimony from Professor Feinstein and Professor Gompers and oral argument on Plaintiffs' motion for class certification and the competing motions to exclude.
On June 12, 2019, the Court denied Defendants' motion to exclude the opinions of Professor Feinstein and denied Plaintiffs' motion to exclude the opinions of Professor Gompers. [Doc. 138] (“Daubert Order”).
The Court now turns to Plaintiffs' Class Certification Motion. Plaintiffs seek certification of a Class consisting of:
All persons who purchased or otherwise acquired The Southern Company common stock between April 25, 2012 and October 30, 2013, inclusive (the “Class Period”), and were damaged thereby.
[Doc. 77-1 at 1] (Plaintiffs' Memorandum in Support of Plaintiffs' Motion for Class Certification).3 Roofers Local No. 149 and Monroe County also seek appointment as Class Representatives and the appointment of Robbins Geller as Class Counsel. Id. For the reasons that follow, Plaintiffs' Class Certification Motion is granted.
III. Legal Standard
“The district court has broad discretion in determining whether to certify a class.” Washington v. Brown & Williamson Tobacco Corp., 959 F.2d 1566, 1569 (11th Cir. 1992). “Before a district court may grant a motion for class certification, a plaintiff seeking to represent a proposed class must establish that the proposed class is adequately and clearly ascertainable.” Little v. T-Mobile USA, Inc., 691 F.3d 1302, 1304 (11th Cir. 2012) (internal quotation marks omitted). Plaintiffs, as the party seeking class certification, must prove they have satisfied the requirements in Rule 23(a) and at least one of the requirements in Rule 23(b). Rule 23(a) requires Plaintiff to show:
(1) the class 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 or defenses of the representative parties are typical of the claims or defenses of the class; and
(4) the representative parties will fairly and adequately protect the interests of the class.
Fed. R. Civ. P. 23(a). These requirements are known as “the prerequisites of numerosity, commonality, typicality, and adequacy of representation.” Gen. Tel. Co. of S.W. v. Falcon, 457 U.S. 147, 156, 102 S.Ct. 2364, 72 L.Ed.2d 740 (1982). Plaintiffs assert that all four prerequisites are satisfied. [Doc. 77-1 at 11–17]. Defendants do not contend otherwise. [Doc. 106].
Once the party seeking certification has shown the requirements of Rule 23(a) are satisfied, the party must show that the putative class meets at least one of the three requirements of Rule 23(b). Here, Plaintiffs seek certification by Rule 23(b)(3), which requires “that the questions of law or fact common to class members predominate over any questions affecting only individual members, and that a class action is superior to other available methods for fairly and efficiently adjudicating the controversy.” Fed. R. Civ. P. 23(b)(3). Those two requirements are known as the “predominance” and “superiority” requirements. Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 615, 117 S.Ct. 2231, 138 L.Ed.2d 689 (1997). Defendants contest only predominance. [Doc. 106 at 8].
Plaintiffs, as the party seeking class certification, must demonstrate by a preponderance of the evidence that the putative class meets the requirements of Rule 23. Thompson v. Jackson, 2018 WL 5993867 (N.D. Ga. 2018) (quoting Brown v. Electrolux Home Prods., Inc., 817 F.3d 1225, 1233 (11th Cir. 2016)). “All else being equal, the presumption is against class certification because class actions are an exception to our constitutional tradition of individual litigation.” Id. Moreover, the Court is required to perform a “rigorous analysis” of the elements of Rule 23(a) and (b). Comcast Corp. v. Behrend, 569 U.S. 27, 35, 133 S.Ct. 1426, 185 L.Ed.2d 515 (2013).
IV. DISCUSSION
A. Rule 23(a)
1. Numerosity
To satisfy the numerosity requirement of Rule 23(a)(1), a movant must show that “the class is so numerous that joinder of all members is impracticable.” Fed. R. Civ. P. 23(a)(1). “ ‘Impracticable’ does not mean ‘impossible’; plaintiffs need only show that it would be extremely difficult or inconvenient to join all members of the class.” In re Domestic Air Transp. Antitrust Litig., 137 F.R.D. 677, 698 (N.D. Ga. 1991) (citing 3 Newberg, Newberg on Class Actions, § 18.03 at 455 (1985)). It is not necessary for the plaintiff to “ ‘allege the exact number․ of purported class members.’ ” In re NetBank, Inc., 259 F.R.D. 656, 664 (N.D. Ga. 2009) (numerosity satisfied for stock listed on the NASDAQ with more than 46 million shares outstanding). The numerosity requirement is “generally assumed to have been met in class action suits involving nationally traded securities.” Zeidman v. J. Ray McDermott & Co., Inc., 651 F.2d 1030, 1039 (5th Cir. 1981); see also Thorpe v. Walter Investment Management, Corp., 2016 WL 4006661, at *6 (S.D. Fla. 2016) (numerosity found for NYSE-traded stock with average weekly trading volume of 3.13 million shares).
Defendants do not dispute numerosity. Further, Plaintiffs have furnished evidence demonstrating that Southern Company's average number of shares outstanding during the Class Period was 872.5 million. [Doc. 77-2, ¶80] (Report on Market Efficiency by Professor Steven P. Feinstein, Ph.D., CFA, dated Sept. 24, 2018 (“Feinstein Rpt.”)). The average weekly trading volume for Southern shares during the Class Period was 22.5 million. Id., ¶50. Thus, the Court finds Plaintiffs have satisfied their burden under the numerosity requirement. In re Vesta Ins. Grp., Inc., Sec. Litig., No. 98-AR-1407-S, 1999 WL 34831475, at *1 (N.D. Ala. Oct. 25, 1999) (finding numerosity satisfied on average weekly trading volume of 345,000 shares on the NYSE).
2. Commonality
Rule 23(a)(2) requires a showing that “questions of law or fact” are common to the class. “The Eleventh Circuit has noted that the Rule 23(a)(2) commonality requirement is a ‘low hurdle.’ ” Thorpe, 2016 WL 4006661, at *6. “ ‘Rule 23 does not require that all the questions of law and fact raised by the dispute be common.’ ” NetBank, 259 F.R.D. at 664 (emphasis in original) (quoting Cox v. Am. Cast Iron Pipe Co., 784 F.2d 1546, 1557 (11th Cir. 1986)). “ ‘ “[E]ven a single [common] question” ’ will do.” Wal-Mart, 564 U.S. at 359, 131 S.Ct. 2541; see also In re HealthSouth Corp. Sec. Litig., 257 F.R.D. 260, 274 (N.D. Ala. 2009) (Commonality's “minimal standard merely requires an identity of some factual or legal matter among members of the class.”). “ ‘Generally, where plaintiffs allege that the action is a result of a unified scheme to defraud investors, the element of commonality is met.’ ” NetBank, 259 F.R.D. at 664 (collecting cases).
Defendants do not dispute commonality. In addition, Plaintiffs allege that all Class members have been harmed as a result of a common course of conduct arising from a common set of material misrepresentations and omissions that Defendants made during the Class Period. Thus, there is a well-defined community of interest in the questions at issue, which include, inter alia:
1. Whether Defendants violated the 1934 Act;
2. Whether Defendants omitted and/or misrepresented material facts;
3. Whether Defendants knew or recklessly disregarded that their statements were false and misleading; and
4. Whether Defendants' statements and/or omissions artificially inflated the price of Southern Company common stock and the extent and appropriate measure of damages.
Each of the above questions focuses on Defendants' conduct and its Class-wide impact, “are susceptible to class-wide proof,” and, thus, “demonstrate[ ] commonality.” In re Internap Network Servs. Corp. Sec., 2012 WL 12878579, at *3 (N.D. Ga. 2012); see also Local 703, I.B. v. Regions Fin. Corp., 282 F.R.D. 607, 612 (N.D. Ala. 2012) (commonality satisfied where “plaintiffs allege a single scheme which violated federal securities law”), vacated in part and aff'd sub nom. in relevant part, Local 703, I.B. of T. Grocery & Food Emps. Welfare Fund v. Regions Fin. Corp., 762 F.3d 1248 (11th Cir. 2014); Cheney v. Cyberguard Corp., 213 F.R.D. 484, 490 (S.D. Fla. 2003) (finding commonality where allegations that defendants “perpetrated a massive fraudulent scheme against investors through uniform misrepresentations and omissions in filings made with the SEC, in press releases, and in other documents”). Thus, Plaintiffs have shown that this case presents common questions of law and fact.
3. Typicality
Rule 23(a)(3) requires that the class representative's claims or defenses be “typical” of the claims or defenses of the putative class. A class “representative's claim is typical if there is a ‘nexus between the class representative's claims or defenses and the common questions of fact or law which unite the class.’ ” In re Wells Real Estate Inv. Tr., Inc. Sec. Litig., No. 1:07-CV-862-CAP, 2009 WL 10688777, at *3 (N.D. Ga. Sept. 16, 2009) (quoting Kornberg v. Carnival Cruise Lines, Inc., 741 F.2d 1332, 1337 (11th Cir. 1984)). A sufficient nexus exists where “ ‘the claims or defenses of the class and the class representative arise from the same event or pattern or practice and are based on the same legal theory.’ ” NetBank, 259 F.R.D. at 665 (quoting Kornberg, 741 F.2d at 1337). “ ‘The typicality requirement may be satisfied despite substantial factual differences․ when there is a strong similarity of legal theories.’ ” Regions, 762 F.3d at 1259-60 (quoting Williams v. Mohawk Indus., Inc., 568 F.3d 1350, 1357 (11th Cir. 2009)).
Here, as with the other Rule 23(a) requirements, Defendants do not dispute typicality. Further, the proposed class representatives – Roofers Local No. 149 and Monroe County – purchased Southern Company common stock during the Class Period, as did all putative Class members. See Plaintiffs' Class Certification Motion; [Doc. 77-5] (“Roofers Local No. 149 Declaration”); [Doc. 77-6 (“Monroe County Declaration”). As a result, Plaintiffs' claims are founded on the same alleged facts and legal theories as the claims of all other Class members – i.e., Defendants' Class Period false statements and omissions and their effect on Southern Company's stock price. Moreover, the injury Plaintiffs suffered is alleged to be the same as the injury suffered by all members of the putative Class. ¶¶225-226; [Doc. 77-2., ¶¶173-179]; Internap, 2012 WL 12878579, at *5 (typicality found where “Plaintiffs, as a class, allege that Defendants issued a number of false and misleading statements․ that artificially inflated the share price of [defendant]'s stock”). Thus, Plaintiffs have shown their claims are typical of those of the Class.
4. Adequacy
Rule 23(a)(4) requires that “the representative parties will fairly and adequately protect the interests of the class.” Courts in the Eleventh Circuit examine a two-prong test for adequacy: “ ‘(1) whether any substantial conflicts of interest exist between the representatives and the class[,] and (2) whether the representatives will adequately prosecute the action.’ ” Dickens v. GC Servs. Ltd. P'ship, 706 F. App'x 529, 535 (11th Cir. 2017) (quoting Valley Drug, 350 F.3d at 1189).
Defendants do not challenge the adequacy requirement, and the Court finds Plaintiffs are adequate. First, there are no substantial or fundamental conflicts of interest between Plaintiffs and the Class. Thorpe, 2016 WL 4006661, at *8 (“Minor conflicts alone will not defeat class representatives' claim to class certification; rather, the conflict must be fundamental, which goes to the specific issues in controversy.”) (citing Valley Drug, 350 F.3d at 1189). Like other potential Class members, Roofers Local No. 149 and Monroe County purchased Southern Company common stock at market prices during the Class Period and were allegedly injured by the same alleged material misrepresentations and omissions that injured all proposed Class members. Plaintiffs' interests in establishing Defendants' liability and maximizing the recovery are aligned with the interests of absent Class members. See Roofers Local No. 149 Declaration and Monroe County Declaration. [Doc. 77-5 and Doc. 77-6].
Furthermore, Roofers Local No. 149 and Monroe County have demonstrated their willingness and ability to serve as class representatives. Plaintiffs have supervised and monitored the progress of the litigation, have participated in discussions with Lead Counsel concerning case developments, have reviewed Court filings, understand their duty to the Class and are committed to vigorously prosecuting this action to maximize recovery for all Class members. Id. In other words, Plaintiffs have demonstrated a willingness to assert and defend the interests of putative Class members. See NetBank, 259 F.R.D. at 666 (“ ‘[A] principal factor in determining the appropriateness of class certification is the forthrightness and vigor with which the representative party can be expected to assert and defend the interests of the members of the class.’ ”) (quoting Kirkpatrick, 827 F.2d at 726). Thus, the Court finds that Roofers Local No. 149 and Monroe County are adequate.
In addition, Rule 23(g)(1) provides that “a court that certifies a class must appoint class counsel.” Plaintiffs request that the Court appoint their chosen Lead Counsel, Robbins Geller, as Class Counsel. In appointing class counsel, the Court considers counsel's work “in identifying or investigating potential claims in the action,” “counsel's experience in handling class actions,” “counsel's knowledge of the applicable law,” and “the resources that counsel will commit to representing the class.” Rule 23(g)(1)(A)(i)-(iv).
Robbins Geller is well qualified to prosecute this case on behalf of Plaintiffs and the other members of the Class. Robbins Geller attorneys have extensive securities litigation experience and have successfully prosecuted numerous securities fraud class actions on behalf of injured investors. See, e.g., Regions, 282 F.R.D. at 616 (“[C]ourts have referred to Plaintiffs' chosen counsel, Robbins, Geller, as ‘one of the most successful law firms in the securities class actions․ in the country.’ ”); [Doc. 77-4] (describing Robbins Geller's extensive history of successful securities fraud matters). Robbins Geller has already undertaken a vigorous prosecution of this action, including conducting an extensive investigation of the claims, defeating Defendants' motion to dismiss and motion for reconsideration of the Court's order on the motion to dismiss, engaging in discovery, and vigorously litigating class certification. Accordingly, Robbins Geller fulfills the requirements of Rule 23(g) and is adequate Class Counsel.
In conclusion, the Court finds that Plaintiffs have satisfied all of the Rule 23(a) requirements.
B. Rule 23(b)(3)
Plaintiffs seek certification pursuant to Rule 23(b)(3), which requires a showing that common questions of fact or law predominate over any individual questions and that maintaining the action as a class action is superior to other available methods for adjudicating the controversy. Rule 23(b)(3).
1. Predominance
Rule 23(b)(3) requires “that the questions of law or fact common to class members predominate over any questions affecting only individual members.” This requirement “tests whether proposed classes are sufficiently cohesive to warrant adjudication by representation.” Amchem Prods. v. Windsor, 521 U.S. 591, 623, 117 S.Ct. 2231, 138 L.Ed.2d 689 (1997). “ ‘It is not necessary that all questions of fact or law be common, but only that some questions are common and that they predominate over individual questions.’ ” Internap, 2012 WL 12878579, at *6. “ ‘[P]redominance is a test readily met’ in [securities fraud] cases such as this.” Id. at *8 (quoting Amchem, 521 U.S. at 625, 117 S.Ct. 2231); see also In re Miller Industries, Inc. Securities Litigation, 186 F.R.D. 680, 688 (N.D. Ga. 1999) (predominance met where “claims of each member of the class arise out of this same set of operative facts”).
“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., 563 U.S. 804, 131 S.Ct. 2179, 180 L.Ed.2d 24 (2011) (“Halliburton I”) (quoting Rule 23(b)(3)). “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.’ ” Id.
Despite the many common issues identified above as part of the Rule 23(a) analysis, Defendants contend that individual issues regarding reliance and damages predominate over any common inquiries.4 The Court addresses these arguments in turn.
a. Reliance
“The reliance element ‘ “ ‘ensures that there is a proper connection between a defendant's misrepresentation and a plaintiff's injury.’ ” ’ ” Halliburton Co. v. Erica P. John Fund, Inc., 573 U.S. 258, 267, 134 S.Ct. 2398, 189 L.Ed.2d 339 (2014) (“Halliburton II”) (quoting Amgen, 568 U.S. at 488, 133 S.Ct. 1184). “The traditional (and most direct) way a plaintiff can demonstrate reliance is by showing that he was aware of a company's statement and engaged in a relevant transaction․ based on that specific misrepresentation.” Halliburton I, 563 U.S. at 810, 131 S.Ct. 2179. In Basic Inc. v. Levinson, 485 U.S. 224, 108 S.Ct. 978, 99 L.Ed.2d 194 (1988), however, the Supreme Court “recognized that requiring such direct proof of reliance ‘would place an unnecessarily unrealistic evidentiary burden on the Rule 10b-5 plaintiff who has traded on an impersonal market’ ” because, even if the plaintiff could show that he was aware of the misrepresentation, he would have to “ ‘show a speculative state of facts, i.e., how he would have acted․ if the misrepresentation had not been made.’ ” Halliburton II, 573 U.S. at 267, 134 S.Ct. 2398 (citing Basic, 485 U.S. at 245, 108 S.Ct. 978). The Supreme Court also recognized that such a requirement would essentially prevent security fraud cases from ever proceeding as class actions; if every plaintiff had to prove direct reliance on a defendant's misrepresentation, individual issues of reliance would predominate over common issues, preventing class certification. Id. at 267-68, 134 S.Ct. 2398 (citing Basic, 485 U.S. at 242, 108 S.Ct. 978).
“To address these concerns, Basic held that securities fraud plaintiffs can in certain circumstances satisfy the reliance element of a Rule 10b-5 action by invoking a rebuttable presumption of reliance, rather than proving direct reliance on a misrepresentation.” Id. at 268, 134 S.Ct. 2398. This rebuttable presumption is based on the “ ‘fraud-on-the-market’ ” theory, “which holds that ‘the market price of shares traded on well-developed markets reflects all publicly available information, and, hence, any material misrepresentations.’ ” Id. (quoting Basic, 485 U.S. at 246, 108 S.Ct. 978). The presumption is that
[a]n investor who buys or sells stock at the price set by the market does so in reliance on the integrity of that price. Because most publicly available information is reflected in market price, an investor's reliance on any public material misrepresentations․ may be presumed for purposes of a Rule 10b-5 action.
Basic, 485 U.S. at 247, 108 S.Ct. 978.
To invoke the Basic presumption, “a plaintiff must prove that: (1) the alleged misrepresentations were publicly known, (2) they were material, (3) the stock traded in an efficient market, and (4) the plaintiff traded the stock between when the misrepresentations were made and when the truth was revealed.” Halliburton II, 573 U.S. at 277-78, 134 S.Ct. 2398 (citing Basic, 485 U.S. at 248, 108 S.Ct. 978; Amgen, 568 U.S. at 471-73, 133 S.Ct. 1184.
The presumption is rebuttable, however, and a defendant may rebut the presumption by “sever[ing] the link between the alleged misrepresentation and either the price received (or paid) by the plaintiff, or his decision to trade at a fair market price.” Basic, 485 U.S. at 248, 108 S.Ct. 978. For example, a defendant could rebut the presumption with evidence that “the․ misrepresentation did not, for whatever reason, actually affect the market price, or that a plaintiff would have bought or sold the stock even had he been aware that the stock's price was tainted by fraud.” Halliburton II, 573 U.S. at 269, 134 S.Ct. 2398.
In this case, Plaintiffs invoke Basic's presumption of reliance in order to demonstrate that common issues of reliance predominate over individual issues. Defendants do not dispute that the alleged misrepresentations were “publicly known” or that Plaintiffs purchased Southern Company common stock between the time the alleged misrepresentations were made and the end of the Class Period. Id. Rather, Defendants argue only that Plaintiffs fail to show the market for Southern Company stock during the Class Period was efficient, which they say precludes Plaintiffs' invocation of the fraud-on-the-market presumption of reliance. Further, Defendants contend that even if Plaintiffs have established that the market for Southern Company stock was efficient during the Class Period such that they can invoke the fraud-on-the-market presumption, Defendants have rebutted the presumption.
The Court addresses Plaintiff's showing of market efficiency and Defendants' attempt to rebut the presumption in turn, below.
(1) Market Efficiency
For purposes of determining market efficiency, the Eleventh Circuit instructs courts to consider the “totality of the circumstances” and confirm the existence of “traditional indicia of efficiency.” Regions, 762 F.3d at 1255. This “traditional indicia” exists “when ‘millions of shares change hands daily and a critical mass of’ investors and/or analysts․ ‘study the available information and influence the stock price through trades and recommendations.’ ” Id. Further indicia of efficiency exist when a company is eligible to file an SEC Form S-3 and has an investor base comprising a wide range of institutional investors. Id. at 1258. Finally, the Eleventh Circuit has noted that “securities trading on national exchanges like the NYSE ‘are often presumed to be traded on an efficient market,’․ precisely because the exchanges are generally populated by stocks that are closely watched by analysts and that trade at a high volume.” Id. at 1257.
The Court finds that all of the “traditional indicia of efficiency” outlined by the Eleventh Circuit in Regions existed with respect to Southern Company stock during the Class Period, and neither Defendants nor their expert dispute this fact. First, “ ‘millions of shares change[d] hands daily’ ” during the Class Period. Id. at 1255. Specifically, the average daily and weekly trading volumes for Southern Company common stock during the Class Period were 4.5 million and 22.5 million shares, respectively. [Doc. 77-2, ¶¶49-50]; see also NetBank, 259 F.R.D. at 670 (average weekly trading volume of 1.36 million indicative of market efficiency). Second, a “ ‘critical mass’ ” of analysts and investors followed Southern Company stock during the Class Period (Regions, 762 F.3d at 1255) – at least 31 different analysts followed Southern Company, there were at least 182 market makers, and at least 1,918 news articles about Southern Company were published during the Class Period. [Doc. 77-2, ¶¶56-58, 68]; see also Internap, 2012 WL 12878579, at *7-*8 (coverage by at least five different broker/dealers indicative of market efficiency). Third, Southern Company was eligible to file an SEC Form S-3 at all times during the Class Period. [Doc. 77-2, ¶73]; see also Internap, 2012 WL 12878579, at *7 n.6 (“ ‘Form S-3 is reserved for companies whose stock is actively traded and widely followed.’ ”). Fourth, at least 1,314 major institutions owned Southern Company stock during the Class Period. [Doc. 77-2, ¶62];5 see also In re HealthSouth Corp. Sec. Litig., 261 F.R.D. 616, 637 (N.D. Ala. 2009) (“HealthSouth II”) (range of 118 to 489 institutional investors indicative of market efficiency). Finally, Southern Company stock traded on the NYSE throughout the Class Period. [Doc. 77-2, ¶¶65-69].6 Because Plaintiffs have shown – and Defendants do not dispute – that the “traditional indicia of efficiency” are satisfied, the Court finds that Plaintiffs have established market efficiency.
In addition, although the Eleventh Circuit has not “adopt[ed] the Cammer factors as the mandatory analytical framework for market efficiency inquiries,” it acknowledges that “some of those factors might prove particularly useful when a District Court considers a stock for which the more traditional indicia of efficiency ․ are not present.” Regions, 762 F.3d at 1255. The five factors set forth in Cammer v. Bloom, 711 F. Supp. 1264 (D.N.J. 1989), are: (1) large trading volume; (2) the existence of a significant analyst coverage; (3) the existence of market makers and arbitrageurs in the security; (4) the eligibility of the company to file an SEC Form S-3 Registration Statement; and (5) demonstration of a cause-and-effect relationship between the announcement of unexpected news and stock price movement. Courts also look to the following three additional factors set forth in Krogman v. Sterritt, 202 F.R.D. 467, 474 (N.D. Tex. 2001): “(1) the capitalization of the company; (2) the bid-ask spread of the stock; and (3) the percentage of stock not held by insiders (the ‘float’).”
The Court finds that all of these factors, some of which overlap with the factors set forth by the Eleventh Circuit in Regions, are satisfied here. Notably, Defendants do not dispute the first four Cammer factors or any of the Krogman factors. In fact, at the evidentiary hearing, Defendants' expert – Professor Gompers – conceded that he did not consider the first four Cammer factors or any of the Krogman factors in analyzing market efficiency. Day 1 Tr. at 207:3-18. Rather, Professor Gompers has argued that the first four Cammer factors are “totally uninformative” of market efficiency. Id. at 209:13-22. The Plaintiffs pointed out that Professor Gompers has been criticized for this opinion in the past:
It may be that many financial economists, including Dr. Gompers, dispute the relevancy of the first four Cammer factors to a determination of market efficiency, but the Cammer factors nonetheless reflect the legal standard for market efficiency․ In fact, this argument has been offered and rejected by Dr. Gompers before. In rejecting the same arguments pursued by Defendants and Dr. Gompers, the district court for the Northern District of Illinois aptly stated, “Defendants rely on factors that are not legally relevant․ Defendants (and their expert) often describe a different conception of an efficient market than is used by the law.”
Willis v. Big Lots, Inc., No. 2:12-cv-604, 2017 WL 1074048, at *4 (S.D. Ohio Mar. 17, 2017).
The Court finds that the Cammer and Krogman factors provide additional evidence of market efficiency, and, therefore, the Court will address each of the Cammer and Krogman factors.
Cammer Factor One – Weekly Trading Volume. “[A]verage 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.” Cammer, 711 F. Supp. at 1286. During the Class Period, an average of 22.5 million Southern Company shares changed hands weekly, which amounted to 2.58% of all Southern Company shares outstanding. [Doc. 77-2, ¶50]. Plaintiffs have satisfied Cammer factor one.
Cammer Factor Two – Analyst Coverage. The existence of “a significant number of securities analysts that followed and reported on a company's stock during the class period” further supports market efficiency. Cammer, 711 F. Supp. at 1286; Regions, 762 F.3d at 1255 (same); NetBank, 259 F.R.D. at 671 (broad analyst coverage “ ‘implies that company reports are ‘closely reviewed by investment professionals, who would in turn make buy/sell recommendations to client investors’ ”). As described above, at least 31 securities analysts covered Southern Company during the Class Period. [Doc. 77-2, ¶¶52-57]; cf. Internap, 2012 WL 12878579, at *7-*8 (coverage by at least five different broker/dealers sufficient); Aranaz v. Catalyst Pharm. Partners Inc., 302 F.R.D. 657, 669 (S.D. Fla. 2014) (at least four analysts sufficient). Plaintiffs have satisfied Cammer factor two.
Cammer Factor Three – Market Makers. The NYSE – one of the most renowned and liquid stock exchanges in the world – employs a Designated Market Maker (“DMM”) for each listed stock. [Doc. 77-2, ¶65]. DMMs are responsible for maintaining a fair and orderly market for registered securities, including the market for Southern Company stock during the Class Period. Id., ¶¶65, 69. In addition, there were 182 market makers for Southern Company stock through the NASDAQ trading platform. Id., ¶68. Trading on the NYSE supports a finding of efficiency. Regions, 762 F.3d at 1257. Plaintiffs have satisfied Cammer factor three.
Cammer Factor Four – SEC Form S-3 Eligibility. The Cammer court found that eligibility to file an SEC Form S-3 supports a finding of market efficiency. Cammer, 711 F. Supp. at 1285;7 see also Regions, 762 F.3d at 1258. A company is eligible to file an SEC Form S-3 if it has filed financial reports with the SEC for 12 consecutive months and has a float of at least $75 million. [Doc. 77-2, ¶¶70-71]. Southern Company – with an average float of $38.9 billion during the Class Period – was eligible to file an SEC Form S-3 throughout the Class Period. Id., ¶73; see also HealthSouth II, 261 F.R.D. at 635 (“The ability to file the abbreviated Form S-3 creates a presumption that the securities trade [is] an efficient market.”). Plaintiffs have satisfied Cammer factor four.
Courts also consider the three market efficiency factors set forth in Krogman. All three of those factors are satisfied here, and Defendants do not contend otherwise. First, Southern Company is one of the largest companies in the United States, with a market capitalization during the Class Period of $39.1 billion. [Doc. 77-2, ¶76]. Second, Southern Company had an average float of 99.5% during the Class Period, meaning that virtually all of Southern Company's stock was available for trading during the Class Period. Id., ¶80. Third, Southern Company stock had an extremely low average bid-ask spread during the Class Period – 0.02%, or $0.01 per share – whereas the average bid-ask spread for all other companies during this period was 0.72%, or $0.10 per share. Id., ¶¶84-85. Plaintiffs have satisfied the Krogman factors.
The Court finds that Plaintiffs have established Cammer factors one through four and all of the Krogman factors, which Defendants do not challenge. Rather, Defendants challenge only Plaintiffs' demonstration of a cause-and-effect relationship between the release of Company-specific news and movements in Southern Company's stock price pursuant to the final Cammer factor – factor five. Although not required for a finding of market efficiency, for the reasons that follow, the Court finds that Plaintiffs have demonstrated a sufficient cause-and-effect relationship in support of Cammer factor five to satisfy this class certification process.
Cammer Factor Five – the Cause-and-Effect Relationship. The fifth Cammer factor considers whether a plaintiff can demonstrate a cause-and-effect relationship between the release of unexpected company news and movements in a company's stock price. Regions, 762 F.3d at 1254 n.2. The Eleventh Circuit in Regions held that the fifth Cammer factor is not a prerequisite to a finding of market efficiency. Id. at 1255-56. “Even the Cammer court itself did not establish such a strict evidentiary burden at the class certification stage.” Id. at 1256. Indeed, the Eleventh Circuit in Regions affirmed the district court's finding of market efficiency even where plaintiffs did not proffer any event study in support of the fifth Cammer factor. Id. at 1256-57. Importantly, the Court is not aware of any case in the Eleventh Circuit, and Defendants cite none, finding a market inefficient where all Cammer/Krogman factors but Cammer factor five were satisfied.
The Eleventh Circuit is not alone in its determination that the fifth Cammer factor is not a prerequisite to a finding of market efficiency – a substantially similar approach has been taken by the First, Second, Third, Fourth, and Fifth Circuits. Waggoner v. Barclays PLC, 875 F.3d 79, 97 (2d Cir. 2017), cert. denied, ––– U.S. ––––, 138 S. Ct. 1702, 200 L.Ed.2d 954 (2018); Carpenters Pension Tr. Fund of St. Louis v. Barclays PLC, 310 F.R.D. 69, 83 (S.D.N.Y. 2015) (citing to First, Second, Third, Fourth, Fifth, and Eleventh Circuit cases holding that the fifth Cammer factor is not necessary). That is because “[r]equiring a plaintiff to submit proof of market reactions – and to do so with an event study – ignores Supreme Court precedent as well as practical considerations.” Barclays, 310 F.R.D. at 84.8
Although not required, Professor Feinstein conducted an event study to empirically demonstrate the cause-and-effect relationship between unexpected company-specific news and movements in Southern Company's stock price. [Doc. 77-2, ¶91]; see also Halliburton II, 573 U.S. at 280, 134 S.Ct. 2398 (endorsing use of event studies in securities cases); FindWhat Inv'r Grp. v. FindWhat.com, 658 F.3d 1282, 1313 & n.31 (11th Cir. 2011) (noting “event studies are a ‘common method’ ” used to assess market efficiency). Using statistical analysis, Professor Feinstein isolated the portion of Southern Company stock movement following each of the seven Class Period earnings announcements that cannot be attributed to market or sector factors, i.e., “the residual stock price movement or ‘residual return.’ ” [Doc. 77-2, ¶93]. Earnings announcement dates are appropriate event dates in event studies investigating a cause-and-effect relationship between the release of company-specific news and company stock price movement.9 Neither Defendants nor Professor Gompers challenge Professor Feinstein's selection of earnings announcements as event dates. Day 1 Tr. at 157:5-14.10
Professor Feinstein examined one-day windows for all seven event dates and found that five of seven dates elicited statistically significant one-day stock price reactions. [Doc. 77-2 ¶168 & Ex. 7].11 These findings satisfy (at this stage of the proceedings) the requirement of a cause-and-effect relationship between the announcement of company-specific news and movements in Southern Company's stock price, which, in combination with the other Cammer and Krogman factors, provides further evidence of market efficiency.12 Additionally, Professor Feinstein conducted a Fisher's Exact Test and found that “the probability of observing five of seven earnings announcements eliciting statistically significant one-day stock price reactions would be only 0.003%.” [Doc 113-2, ¶92]. Put differently, observing significant reactions on five of the seven event dates rejects “to the 99.997% confidence level” the hypothesis that Southern Company stock did not react to new, value-relevant information. Id.13 The Court finds that Professor Feinstein's empirical analysis analyzing exclusively one-day event windows is strong evidence of market efficiency.
Further, for two earnings announcements toward the end of the Class Period – July 31 and October 30, 2013 – Professor Feinstein stated that he examined and reported the results for both one-day and two-day event windows in light of the particular facts and circumstances surrounding those events. [Doc. 77-2, ¶¶156-167 & Ex. 7].14 Professor Feinstein explained why the nature and timing of the news released on July 31, 2013 supported investigating the second day before concluding that there was no reaction to the July 31st event. First, Professor Feinstein explained that the July 31, 2013, news was complex – it was corrective information concerning the alleged fraud that contradicted earlier Company representations and was released near the end of the Class Period, which required “analysts and investors to conduct substantial reevaluation of the Company.” [Doc. 113-2, ¶¶75-77]. Second, the majority of analyst reports concerning the July 31, 2013, event were issued after the close of market on July 31, 2013, and continued on to August 1, 2013, demonstrating that the market required additional time to incorporate the news released late in the trading day on July 31, 2013. Id., ¶¶80, 83.15 Third, there was an earnings conference call with analysts and investors that began at 1:00 p.m. Eastern and ended at approximately 2:30 p.m. Eastern on July 31, 2013, just 90 minutes before the close of the market, during which Defendants elaborated on the disappointing financial results and answered questions from analysts. Id., ¶84. Fourth, during the 1:00 p.m. Eastern earnings call, analysts learned for the first time that Southern Company would need to issue up to $700 million in equity to protect the balance sheet as a direct result of mounting Kemper Plant costs, causing nearly all analysts who issued post-market close reports on July 31, 2013, and August 1, 2013, to report on this surprising, negative, and new information and cite it as a reason to reduce their estimates going forward. Id., ¶87.16 Considering the two-day window for the July 31, 2013 event date, six of seven event dates were followed by statistically significant price reactions, and the one date that was not – the “mixed” January 30, 2013 announcement – was followed by an appropriately non-significant price reaction. The Court finds that Professor Feinstein's analysis of one- and two-day event windows, in addition to his other, non-empirical findings, provides further evidence of market efficiency.
The Court rejects Defendants' argument that Professor Feinstein's event study failed to establish sufficiently an evidence of a cause-and-effect relationship. The Court notes at the outset that Professor Gompers does not opine that the market for Southern Company stock was inefficient during the Class Period. Day 1 Tr. at 206:20-207:2; cf. Scientific-Atlanta, 571 F. Supp. 2d at 1340 (noting that where defendant's market efficiency expert only criticizes plaintiff's expert's analysis and “does not opine that the market․ was not efficient․, the Court affords his affidavit little weight”). Rather, Professor Gompers proffers a different event study model, using a different industry sector index 17 and analyzing only one-day event windows, that finds that three of seven event dates were followed by statistically significant one-day price reactions. [Doc. 106-2] (Expert Report of Paul A. Gompers, Ph.D., dated Jan. 11, 2019 (“Gompers Rpt.”)).
While the Court declines to “engage in the parties' battle of the experts” with respect to Cammer factor five (NetBank, 259 F.R.D. at 670 & n.8, 674), the Court notes that Professor Gompers' event study results themselves support a finding of market efficiency.18 And, in any event, event studies are not required in this Circuit for a finding of market efficiency.19 Despite the non-dispositive nature of the fifth Cammer factor and the Court's disinclination to engage in a battle of the experts at this stage, the Court has considered Defendants' arguments with respect to the sector index and event window and finds that Professor Feinstein's selections were appropriate in this case.
(i) Sector Index
As noted above, the sector index in an event study model “control[s] for the effects of market-wide and industry news on the company's stock price, so that one can isolate the effect of company-specific news.” [Doc. 113-2, ¶99]. In this case, Professor Feinstein employed his standard methodology of selecting the index chosen by the target company in its SEC filings for purposes of complying with SEC Regulation S-K, which is an SEC regulation that requires publicly traded companies to identify an industry sector index so that readers can compare the target company's stock performance against the performance of peer companies in the target company's sector. Id., ¶¶100-101. Pursuant to SEC Regulation S-K, Southern Company elected to utilize throughout the Class Period a “published industry or line-of-business index.” 17 C.F.R. § 229.201. Specifically, the Company chose an index published by Standard & Poor's – the Standard & Poor's Electric Utility Index (“S&P EUI”) – as the representative sector index against which investors should compare Southern Company's stock performance. [Doc. 113-2, ¶102].20 Standard & Poor's, a global financial services company, “is recognized as the industry leader in the creation and production of sector indices.” Day 1 Tr. at 75:9-16. Standard & Poor's included in the S&P EUI only those companies deemed to be “electric utility” companies, such as Southern Company. Id. at 78:7-79:6.21
Professor Gompers, on the other hand, created and tested six different sector indices in search of the index that had the highest “R value” or tightest “fit” vis-à-vis Southern Company's stock price movements. [Doc 106-2 ¶¶56, 62-63]. Professor Gompers settled on an index he created by including companies identified by Southern Company as “Custom Peer Group” companies for purposes of executive compensation benchmarking purposes in its 2013 and 2014 proxy statements. Id., ¶63 n.46.
As the Court explained in its Daubert Order, “two experts can disagree on the index which best represents the company at issue when conducting an event study.” [Doc. 138] Daubert Order at 15; accord Carpenters Health & Welfare Fund v. Coca-Cola Co., No. 1:00-CV-2838-WBH, 2008 WL 4737173, at *2 (N.D. Ga. Mar. 14, 2008) (rejecting challenge to Professor Feinstein's sector index because “[i]n any complex statistical analysis, there is the potential for reasonable minds to disagree regarding these choices”). Nonetheless, “[a]fter reviewing the information regarding the index chosen by Dr. Feinstein, the Court is unable to see where an index reported to the SEC by Defendant Southern would be an inappropriate choice.” [Doc. 138 at 10]. The Court finds that Professor Feinstein's selection of the S&P EUI and the use of that index in event study analysis for this case were appropriate.
For a number of reasons, including that the Defendants' arguments concerning which is the best sector index are better left for trial, the Court finds no reason to jettison the S&P EUI in favor of Professor Gompers' “Custom Peer Group” index. First, Professor Gompers claims that his index is better because it has better statistical “fit.” [Doc. 106-2, ¶63]. Even if that is true, the trier of fact at trial will be tasked with weighing each expert's index selection and deciding which index best matches Southern Company's sector. [Doc. 138 at 10].
Second, Professor Gompers cited no authority supporting his opinion that a researcher should create and test multiple indices and use an index with the tightest fit. Indeed, Professor Gompers' two authorities on this point, which he cites to support only the definition of statistical fit, do not state that one should search for and use an index with the tightest statistical fit. To the contrary, those authorities make clear that using an index with the tightest fit or highest R value is not necessarily preferable.22
Third, Professor Gompers himself has previously opined that choosing an index based on statistical fit is improper data-mining and not accepted by the academic literature. See [Doc. 113-2, ¶141]. Specifically, in In re Northfield Labs, Sec, Litig., Professor Gompers opined that plaintiff's expert was engaged “ ‘in what statisticians call “data-mining,” which in this case means selectively choosing a set of companies that provides the best fit for his model. This data mining approach is not accepted in the academic literature. Rather the accepted and standard way to select competitors is to use an a priori selection criterion.’ ” Id. In Northfield, Professor Gompers cited to MacKinlay, et al., for the proposition that “the most effective means of reducing the impact of over-fitting and data-snooping is to impose some discipline on the specification search by a priori considerations.” “The Econometrics of Financial Markets,” by Craig A. MacKinlay, John Y. Campbell, and Andrew W. Lo (Princeton University Press, 1997), p. 524.
Fourth, whereas the S&P EUI was an index used by Southern Company during the Class Period, Professor Gompers' so-called “Custom Peer Group” was not. [Doc. 113-2, ¶¶117, 131].23 Rather, the “Custom Peer Group,” as referenced in Southern Company's proxy statements, was a list of companies that Southern Company identified as compensation benchmarks – in other words, companies that Southern Company benchmarked to determine how to pay its executives. [Doc. 113-2, ¶131]. But companies choose compensation benchmark comparators for myriad reasons that may have nothing to do with the industry sector. For example, companies choose compensation benchmark comparators that have similar management structure, number of employees, geographic locations, or market capitalization regardless of whether those comparators are in the same industry sector. Day 1 Tr. at 79:11-18; ECF No. 139 (Transcript of May 22, 2019 Proceedings Before The Honorable William M. Ray, II (“Day 2 Tr.”)) at 29:13-30:7. Indeed, the S&P EUI was comprised only of companies deemed to be “electric utility” companies and excluded companies deemed to be “multi-utility” companies, which are companies that provide other, non-electric utility services, such as water and natural gas. Day 1 Tr. at 78:7-79:6. Several of the companies included in Professor Gompers' “Custom Peer Group” index, however, were “multi-utility” companies. Id. And Southern Company indicated that its compensation benchmark companies were selected because they were “ ‘believed to be most similar to the Company in both business model and investors,’ ” not because those companies were in the same industry sector. [Doc 113-2, ¶¶135-136]; see also Day 2 Tr. at 29:13-30:7. To be sure, companies can have the same “business model and investors” as Southern Company but not be in the same industry sector. Day 2 Tr. at 29:13-30:7 (explaining that compensation peers can be selected because, for example, they have similar management structure or are in the same geographic region, irrespective of whether they operate in the same industry sector).
Finally, the Court finds compelling the fact that Southern Company, throughout the Class Period, represented in its SEC filings that the S&P EUI was the industry sector against which investors should compare Southern Company's stock performance. Professor Gompers' post-hoc attempt to create new indices with tighter statistical fit, despite the fact that the Company never did so and never described any deficiencies with the S&P EUI (both of which SEC Regulation S-K expressly invites companies to do) compels the Court to conclude that use of the S&P EUI index is appropriate in this action.
(ii) Event Window Duration
The “event window” in an event study model refers to the time period over which the stock price of the target company will be examined vis-à-vis the event under examination. [Doc. 77-2., ¶159]. As explained above, Professor Feinstein examined and presented results using one-day event windows for all seven earnings announcements, meaning that he examined the stock price reaction on the day of the earnings announcement only. Id., ¶168 & Ex. 7. For two dates toward the end of the Class Period – July 31 and October 30, 2013 – Professor Feinstein analyzed and presented the results using two-day event windows, meaning that he examined the earnings announcement date and the date following the earnings announcement. Id., ¶¶156-167. Professor Feinstein explained why the facts and circumstances warranted examination of the second day for these event dates. See supra at 386–87.
As the Court explained in its Daubert Order, “courts recognize the particular circumstances of an event may dictate the window used by an expert when conducting event studies.” [Doc. 138 at 8]. For a number of reasons, the Court declines to limit its consideration of the experts' event study results to one-day event windows only.
First, using exclusively one-day event windows does not change Professor Feinstein's conclusion concerning market efficiency. [Doc. 113-2, ¶¶91-95]. As a result, the debate about event windows in this case is largely academic.
Second, there are many cases that find that multi-day event windows are appropriate for event study analysis in securities fraud class actions. See, e.g., In re Xcelera.com Sec. Litig., 430 F.3d 503, 513 n.11 (1st Cir. 2005) (rejecting argument that a two-day event window is inconsistent with an efficient market); In re DVI, Inc. Sec. Litig., 639 F.3d 623, 635 (3rd Cir. 2011) (“That some information took two days to affect the price does not undermine a finding of efficiency.”), abrogated on other grounds by Amgen Inc. v. Connecticut Retirement Plans and Trust Funds, 568 U.S. 455, 133 S. Ct. 1184, 185 L.Ed.2d 308 (2013); In re Vivendi Universal, S.A., Sec. Litig., 634 F. Supp. 2d 352, 372 (S.D.N.Y. 2009) (using a three-day window for analysis); In re Diamond Foods, Inc., 295 F.R.D. 240, 249 (N.D. Cal. 2013) (citing defense expert's assertion that a “proper test of market efficiency” required analyzing whether the market price reacted to new information over more than one trading day in order to “assess[ ] whether there are delayed price responses to the events of interest”); Fogarazzo v. Lehman Bros., 263 F.R.D. 90, 104 (S.D.N.Y. 2009) (accepting event study with three-day event window and rejecting argument that this rule only applied “where the timing of discrete events was [not] ascertainable”); Aranaz, 302 F.R.D. at 669 (citing significant two-day stock decline as “strong empirical evidence of market efficiency”).
In another securities fraud case, Carpenters Pension Tr. Fund of St. Louis v. Barclays PLC, the court explicitly considered and rejected Professor Gompers' exact same argument against two-day windows that he offers in this case, finding that “[a] two-to three-day window is common in event studies” because “it is standard for experts to utilize an event window including both the day of the event and the day following an event.” 310 F.R.D. at 96. The Barclays court noted that Professor Gompers “agree[d] that event studies often use a two-day window, the date of the announcement and the day after.” Id. at 96 n.183.
Third, and more broadly, the Supreme Court has expressly refused to “adopt any particular theory of how quickly and completely publicly available information is reflected in market price.” Basic, 485 U.S. at 248 n.28, 108 S.Ct. 978. Following its decision in Basic, the Supreme Court in Halliburton II reiterated that it would not “enter the fray” of academic debates about the speed at which information is impounded into a stock price, and reconfirmed that market efficiency is based on “the fairly modest premise that ‘market professionals generally consider most publicly announced material statements about companies, thereby affecting stock market prices.’ ” 573 U.S. at 271-72, 134 S.Ct. 2398 (quoting Basic, 485 U.S. at 247 n.24, 108 S.Ct. 978).
Fourth, academic literature supports the use of two-day (or more) event windows. For example, MacKinlay explains that two-day event windows are appropriate when analyzing earnings announcements.24 Professor Feinstein cited multiple other legal and academic papers endorsing the use of multi-day event windows and explaining that price reactions to news sometimes occur over two or more trading days, including a paper by David I. Tabak, who explains that “[i]n securities fraud cases, many experts have adopted the convention of looking at one-day, two-day, or five-day periods following an announcement.”25 And although Defendants and Professor Gompers cite academic literature for the proposition that stock price reactions generally begin quickly, this does not necessarily mean that stock price reactions always end quickly. Rather, as Professor Feinstein explained, “[i]t is not settled science that an efficient market price reaction is over within a matter of minutes,” and “the articles [Professor Gompers cites] do not conclude that the stock price reaction in an efficient market necessarily ends quickly and cannot continue on the day after an earnings announcement.” [Doc. 113-2, ¶¶60, 65].
Fifth, the Court finds, in accordance with the academic literature, that when the particular facts and circumstances justify investigating multi-day event windows, such analysis is appropriate. In this case, Professor Feinstein explained that the news released on July 31, 2013, was (1) complex and contradicted previous misrepresentations, (2) released late in the trading day, (3) expanded upon with further clarification and explanation, including that the Company would have to issue $700 million in equity to address Kemper cost overruns just 90 minutes before the market closed,26 and (4) covered by analysts, the majority of whom issued reports after the market closed on July 31, 2013 and on August 1, 2013. See supra at 386–87. Given the factual scenario surrounding the news on July 31, 2013, the Court finds Professor Feinstein's investigation of the price reaction that occurred on August 1, 2013, to be appropriate.
Finally, the Court rejects Defendants' suggestion that Professor Feinstein's analysis should be disregarded because he did not analyze only one-day windows or only two-day windows. The Court is unaware of any rule, and Defendants cite none, requiring that one must commit at the outset, before analyzing the data, to investigate only one-day or only two-day event windows. Professor Feinstein testified that event study analysis is “an investigation, and the results might compel you to look at a larger window.” Day 1 Tr. at 138:4-11; see also id. at 143:7-10 (“[T]o commit to one-day windows only and disregard the possibility that the reaction will be evident on the second day is not reliable investigative research.”); id. at 143:17-24 (“Because the literature says that the reaction might be on the second day, it's imperative to look on the second day before concluding that there was no reaction.”).27
For the foregoing reasons, the Court finds that Professor Feinstein's selection of the S&P EUI and his analysis of one- and two-day event windows was appropriate and credits his event study findings. In combination with Professor Feinstein's other market efficiency findings, including those related to the Eleventh Circuit's “indicia of efficiency” and the other Cammer and Krogman factors, the Courts finds that Plaintiffs have established by a preponderance of the evidence that the market for Southern Company stock was efficient, such that Plaintiffs are entitled to the fraud-on-the-market presumption of reliance.
(2) Price Impact
“[O]nce a plaintiff shows entitlement to a presumption of reliance, the defendant is burdened with the daunting task of proving that the publicly known statement had no price impact.” Aranaz, 302 F.R.D. at 673; Thorpe, 2016 WL 4006661, at *13 (“Defendants must demonstrate that the corrective disclosures played no part in the decline in the Company's share price.”). To meet their burden, Defendants must “show[ ] that the alleged misrepresentation[s] did not actually affect the stock's market price.” Halliburton II, 573 U.S. at 282, 134 S.Ct. 2398; see also id. at 284, 134 S.Ct. 2398 (Ginsburg, J., concurring) (“[I]t is incumbent upon the defendant to show the absence of price impact.”). In other words, Defendants must show that “the alleged misstatements [had] no price impact whatsoever” to rebut the presumption. Emergent Biosolutions, 322 F. Supp. 3d at 687.
To prove a lack of price impact, Defendants and Professor Gompers analyzed the stock price reaction following the five alleged corrective disclosures, which occurred on April 24, July 1, July 31, October 2, and October 30, 2013.28 Defendants cited to Professor Gompers' event study analysis of these five disclosures, focusing on whether the disclosures were followed by statistically significant stock price declines. [Doc. 106] (“Defendants' Class Certification Opposition”) at 17-18. According to Professor Gompers' model, only the April 24, 2013, disclosure was followed by a statistically significant stock price decline. Id. On that basis, Defendants concede that they cannot prove an absence of price impact for that date, but ask the Court to hold as a matter of law that the remaining four corrective disclosures should be dismissed from the case. Id.
The Court finds that Defendants have failed to prove an absence of price impact. As an initial matter, as explained in the Daubert Order and above, the Court finds that Professor Feinstein's event study model, including his index selection and use of one- and two-day event windows, was appropriate. Professor Feinstein's model demonstrates that four of the five alleged corrective disclosures – including the first corrective disclosure on April 24, 2013, and the last corrective disclosure on October 30, 2013 – were followed by statistically significant price declines. [Doc. 113-2, ¶¶176-177 and p. 110]. Thus, because Defendants concede that a statistically significant price decline following an alleged corrective disclosure means one cannot rule out price impact, Professor Feinstein's model demonstrates that Defendants cannot prove an absence of price impact during the Class Period.
Moreover, even if the Court were to rely on Professor Gompers' model, Defendants still have not rebutted the presumption because their price impact arguments rely on a statistical fallacy. Contrary to Defendants' argument, the existence of non-statistically-significant stock price declines does not prove the absence of price impact. “[I]f a price movement is not statistically significant, one cannot rule out random volatility, but neither can one rule out information as the cause of the price movement or at least a contributing factor.” [Doc. 113-2, ¶165]. Thus, “Professor Gompers' failure to find statistically significant stock price declines on the other corrective disclosure event dates during the Class Period proves nothing,” because a “failure to find statistical significance does not prove that information had no role in the observed stock price adjustment.” Id., ¶29. Rather, “[n]on-significance means indeterminate with respect to finding the cause of a stock price movement; it does not mean that there was no decline or that the decline was necessarily caused by factors other than the corrective disclosure.” Id.
An event study tests whether one can reject a null hypothesis. For price impact purposes, the null hypothesis under examination is that the stock price of the subject company was not impacted by the alleged misrepresentations. It is axiomatic that “failure to rebut the null hypothesis does not necessarily mean that a misrepresentation had no price impact.” Jill E. Fisch et al., The Logic and Limits of Event Studies in Securities Fraud Litigation, 96 Tex. L. Rev. 553, 611 (2018). Rather,
[a]n event date's excess returns might be in the direction consistent with the plaintiff's allegations but be too small to be statistically significant at a significance level as demanding as 5%. Failure to demonstrate this level of statistical insignificance does not prove the null hypothesis, however; rather, such failure simply implies that one does not reject the null hypothesis at that significance level. That is, the standard event study does not show the information did not affect stock price; it just shows that the information did not have a statistically significant effect at the 5% level.
Id. As Brav & Heaton explain, “[c]ourts err because of their mistaken premise that statistical insignificance indicates the probable absence of a price impact.” Alon Brav & J.B. Heaton, Event Studies in Securities Litigation: Low Power, Confounding Effects, and Bias, 93 Wash. U. L. Rev. 583, 587 (2015); see also id. at 602 (“[W]hile a statistically significant reaction to a firm-specific news event is evidence [of price impact]․, the converse is not true – the failure of the price to react so extremely as to be two standard deviations from average does not establish that the market is inefficient.”);29 “Statistical Techniques in Business and Economics,” by Robert D. Mason, Douglas A. Lind & William G. Marchal, 10th Edition, Irwin McGraw-Hill, 1999, p. 307 (“We should emphasize that if the null hypothesis [H0] is not rejected, based on the sample data, we cannot say that the null hypothesis is true. To put it another way, failing to reject the null hypothesis does not prove that [the null hypothesis] is true, it means we have failed to disprove [the null hypothesis].”).
Thus, Professor Gompers conceded at the evidentiary hearing that a non-statistically significant decline does not prove the null hypothesis to be true – in other words, failure to observe statistically significant declines does not prove that Southern Company's stock price was not impacted by the alleged misrepresentations. Day 1 Tr. at 247:9-14. Instead, Professor Gompers argued that, “[a]s an academic if you have no evidence that there's price impact[,] you conclude there's no price impact.” Id. at 247:21-23. However, the Court declines to find an absence of price impact simply because Professor Gompers found the price decline to not be statistically significant.30
In recognition of this basic truism of statistics, courts routinely reject the argument that a non-statistically significant stock price decline proves an absence of price impact. For example, the court in Rooney v. EZCORP, Inc. rejected “Defendants' attempt to rebut the Basic presumption” by pointing to the fact that only two of the four alleged corrective disclosures were followed by a statistically significant stock price decline, finding the argument “flawed from a statistical perspective.” Rooney v. EZCORP, Inc., 330 F.R.D. 439, 450 (W.D. Tex. 2019). The court explained:
Defendants suggest the lack of a statistically significant price adjustment following a corrective disclosure shows that whatever price adjustment has occurred must be due to “random chance” rather than a predicate misrepresentation. But that is not how hypothesis testing works. A statistically significant price adjustment following a corrective disclosure is evidence the original misrepresentation did, in fact, affect the stock price. The converse, however, is not true – the absence of a statistically significant price adjustment does not show the stock price was unaffected by the misrepresentation.
Id.31 Similarly, the Third Circuit Court of Appeals recently held that although a non-statistically significant price decline, without more, may not “demonstrate a price impact,” neither is it “necessarily proof of the opposite.” Vizirgianakis v. Aeterna Zentaris, Inc., No. 18-2474, 775 Fed.Appx. 51, 53, 2019 WL 2305491, at *2 (3d Cir. May 30, 2019). The weight of authority is in accord. See, e.g., City of Sterling Heights Gen. Emps.' Ret. Sys. v. Prudential Fin., Inc., No. 12-5275, 2015 WL 5097883, at *13 n.8 (D.N.J. Aug. 31, 2015) (“[I]t also does not necessarily follow from the mere absence of a statistically significant change in the stock price that there was no price impact.”); DFC Glob., 2016 WL 4138613, at *14 (same); Pirnik v. Fiat Chrysler Autos., N.V., 327 F.R.D. 38, 46 (S.D.N.Y. 2018) (holding that non-statistically significant stock price decline following corrective disclosure “does not prove the absence of price impact”) (emphasis in original); Barclays, 310 F.R.D. at 95 (“The failure of an event study to find price movement does not prove lack of price impact with scientific certainty.”); Bing Li v. Aeterna Zentaris, Inc., 324 F.R.D. 331 (D.N.J. 2018) (same), aff'd sub nom. Vizirgianakis v. Aeterna Zentaris, Inc., 775 Fed.Appx. 51, 2019 WL 2305491. Defendants must prove that the corrective disclosures “had no negative impact on the price of” Southern Company stock. Aranaz, 302 F.R.D. at 672; see also Basic, 485 U.S. at 248, 108 S.Ct. 978 (defendants must “sever[ ] the link between the alleged misrepresentation and either the price received (or paid) by the plaintiff” in order to rebut the establish no price impact). A non-statistically significant decline simply does not “sever the link” between the alleged misrepresentations and corrective disclosures.32
In any event, the Court finds that Defendants have failed to rebut the presumptions because they do not even argue that there was “no price impact whatsoever.” Emergent Biosolutions, 322 F. Supp. 3d at 687. Rather, they concede that they cannot rule out price impact because the stock price decline following at least one Class Period corrective disclosure – the April 24, 2013 corrective disclosure – was statistically significant [Doc. 106 at 17]. This concession dooms Defendants' attempt to rebut the presumption of reliance because the inquiry is whether Defendants have proven a complete lack of price impact during the Class Period, not whether the stock price decline following individual corrective disclosures was caused by the alleged misrepresentations, which is a loss causation analysis not appropriate at this stage.
For this reason, numerous courts addressing class certification have refused to shorten class periods by dismissing subsequent corrective disclosures where some but not all of the stock price declines following the alleged corrective disclosures were statistically significant. Instead, these courts found that the question of what caused the stock price to decline is an ultimate merits question for which plaintiffs bear the burden at trial, not at class certification. See Halliburton I, 563 U.S. at 813, 131 S.Ct. 2179 (plaintiffs need not “show loss causation as a condition of obtaining class certification”).
For example, the district court in Regions, on remand from the Eleventh Circuit to consider price impact post-Halliburton II, declined to find no price impact with respect to a corrective disclosure simply because the price decline was not statistically significant. Local 703, I.B. of T. Grocery and Food Employees Welfare Fund v. Regions Financial Corp., 2014 WL 6661918, at *8-*9 (N.D. Ala. 2014). Rather, the Court noted, Regions' price decline following the corrective disclosure and analysts' commentary attributing the decline to the negative announcement was, “of course, evidence of price impact.” Id. at *7. The court held that “[w]hether this tumble was due to defendants' corrective disclosures․ or due to the overall market conditions on that day, is an ultimate question in this action, and properly reserved for a jury to decide.” Id. at *8. Similarly, the court in KBC Asset Mgmt. NV v. 3D Sys. Corp. declined to shorten the class period where only the first of two alleged corrective disclosures was followed by a statistically significant stock price decline. KBC Asset Mgmt. NV v. 3D Sys. Corp., No. 0:15-2393-MGL, 2017 WL 4297450, at *7-*8 (D.S.C. Sept. 28, 2017). The 3D Sys. court found that although the stock price decline following the second corrective disclosure was not statistically significant, “[o]n the record before it, the Court is unable to say Defendants have presented evidence sufficient to convince it there was no price impact associated with” that disclosure because “[w]hether the stock price [decline] was caused by alleged misrepresentations or some other factor is for now an open question.” Id. at *8. Similarly, having found price impact for at least one corrective disclosure, the court in Grae v. Corr. Corp. of Am. rejected defendants' argument that the class period should be shortened because later corrective disclosures were not followed by statistically significant price declines: “Because [defendant] has failed to rebut the Basic presumption regardless, that issue [i.e., whether later corrective disclosures evidence price impact] should [be] left for the merits stage of litigation.” Grae v. Corr. Corp. of Am., 330 F.R.D. 481, 500 n.8 (M.D. Tenn. 2019). And in Rooney, the court rejected defendants' argument that the class period should be shortened because only two of the four alleged corrective disclosures were followed by statistically significant price declines. Rooney, 330 F.R.D. at 448-51.
Further, Professor Feinstein described the stock price declines that occurred following each of the five alleged corrective disclosures, which are indisputable, and analyzed the market's reaction to each of those disclosures. [Doc.113-2, ¶¶178-224].33 Although Plaintiffs need not prove loss causation at this stage (Halliburton I, 563 U.S. at 813, 131 S.Ct. 2179), Professor Feinstein noted that following each of the five disclosure dates, analysts highlighted the negative news and incorporated it into their downward revisions of Southern Company. [Doc. 113-2, ¶182 (April 24, 2013), ¶187 (July 1, 2013), ¶198 (July 31, 2013), ¶207 (October 2, 2013), & ¶218 (October 30, 2013)]. Further, Professor Feinstein testified that he analyzed the analyst reports following each corrective disclosure and found that “[a]nalysts attributed the declines to the negative news.” Day 1 Tr. at 100:21-25. As the court in Regions found when considering price impact, the existence of a price decline and analyst commentary highlighting the negative news is, “of course․ evidence of price impact.” Regions, 2014 WL 6661918, at *7.
Because Southern Company's stock price suffered negative residual returns following each of the alleged corrective disclosures, Defendants' citation to non-significance “merely suggest[s] that another factor also contributed to an impact on a security's price.” Waggoner, 875 F.3d at 105. That “does not establish that the fraudulent conduct complained of did not also impact the price of the security.” Id.; see also Prudential, 2015 WL 5097883, at *12 (merely “introduc[ing] evidence raising a triable issue of fact as to whether there was a price impact” does not rebut the presumption); DFC Glob., 2016 WL 4138613, at *14 (“Because a defendant's burden of proving a lack of price impact is ‘daunting,’ simply ‘pointing to other potential causes for a stock price change following a corrective disclosure is therefore not enough to rebut the Basic presumption.’ ”); Thorpe, 2016 WL 4006661, at *14 (defendants must “conclusively establish that there is no connection between the price decline and the alleged misrepresentation”).34
The Court finds that Defendants have failed to prove an absence of price impact. The remaining questions – what caused Southern Company's stock price to decline following each of the corrective disclosures (i.e., loss causation) and how much inflation was dissipated as a result of those disclosures (i.e., damages) – are ultimate questions for the trier of fact on the merits.
b. Damages
Defendants contend that Plaintiffs have failed to articulate a common damages methodology capable of calculating class-wide damages. First, Defendants argue that Plaintiffs have failed to proffer a damages model tailored to the needs of the case – specifically, “how Plaintiffs would calculate damages allegedly suffered by investors who bought stock at different times; prices; and levels of inflation.” [Doc. 106 at 25]. Second, Defendants argue that Plaintiffs' damages model is inconsistent with their alleged theory of liability, violating the Supreme Court's holding in Comcast Corp. v. Behrend, 569 U.S. 27, 133 S. Ct. 1426, 185 L.Ed.2d 515 (2013). Notably, however, Professor Gompers does not opine that damages cannot be calculated on a class-wide basis; he merely claims that Professor Feinstein fails to propose a model that does so. Compare Scientific-Atlanta, 571 F. Supp. 2d at 1340 (noting that where defendants' expert only criticizes plaintiffs' expert, “the Court affords his affidavit little weight”).
Plaintiffs, on the other hand, argue that the damages model proposed by Professor Feinstein is capable of measuring damages on a class-wide basis, that the model is consistent with Plaintiffs' theory of liability, and that Defendants' arguments are based on a mischaracterization of Plaintiffs' theory of liability in this case. [Doc. 113 at 10-15] (Plaintiffs' Reply Memorandum of Law in Support of Motion for Class Certification). For the reasons that follow, the Court finds that Defendants' damages criticisms do not defeat the predominance requirement of Rule 23(b)(3).
It is axiomatic that individualized damages calculations are generally insufficient to foreclose class certification, and particularly so where the central liability question is common to each class member. See, e.g., Carriuolo v. Gen. Motors Co., 823 F.3d 977, 988 (11th Cir. 2016); In re Delta/AirTran Baggage Fee Antitrust Litig., 317 F.R.D. 675, 686 (N.D. Ga. 2016). Indeed, nothing in Rule 23(b)(3) requires Plaintiffs to prove predominance separately as to both liability and damages; nor did the Supreme Court set forth such a standard in Comcast. See Carriuolo, 823 F.3d at 988; Roach v. T.L. Cannon Corp., 778 F.3d 401, 407 (2d Cir. 2015) (“Comcast․ did not hold that proponents of class certification must rely upon a classwide damages model to demonstrate predominance.”); Butler v. Sears, Roebuck & Co., 727 F.3d 796, 801 (7th Cir. 2013) (“If the issues of liability are genuinely common issues, and the damages of individual class members can be readily determined in individual hearings, in settlement negotiations, or by creation of subclasses, the fact that damages are not identical across all class members should not preclude class certification.”); Pulaski & Middleman, LLC v. Google, Inc., 802 F.3d 979, 987 (9th Cir. 2015) (same).
Thus, at the class certification stage, while the Court must determine whether Plaintiffs have articulated a damages model capable of calculating damages stemming from the Defendants' actions on a class-wide basis, a determination in the negative is not necessarily fatal to class certification.35 Rather, if damages cannot be ascertained by a methodology applicable to all class members, the Court must then consider whether questions of individual damages overwhelm the questions of liability that are subject to common proof. Nevertheless, because the Court finds that Plaintiffs have proffered a damages model consistent with their theory of liability and capable of calculating damages on a class-wide and per share basis, the Court does not need to reach the latter inquiry.
Plaintiffs, through Professor Feinstein, propose a three-step method for calculating damages on a per share basis in this action. [Doc. 77-2, ¶¶173-179]. First, Professor Feinstein will use “valuation tools, which would include event study analysis” to measure artificial inflation in the stock due to the alleged misrepresentations and omissions, as well as the corresponding dissipation caused by corrective disclosures. Id., ¶177(i). Second, Professor Feinstein will create an “inflation ribbon,” which is a “time series of the difference between a stock's actual price observed in the marketplace and the estimated price that the stock would have traded at each day had there been full disclosure from the outset of the Class Period,” to determine the extent of artificial inflation in the stock's price on each day in the Class Period. Id., ¶177(ii). Third, using the inflation ribbon, Professor Feinstein will calculate the damages for each Class member by determining “the difference between the inflation on the date the shares were purchased and the inflation on the date those same shares were subsequently sold, excluding any inflation dissipation caused by factors other than corrective disclosure,” while limiting per share damages “to be no greater than the decline in share price over the holding period.” Id., ¶177(iii); Day 1 Tr. at 121:7-10, 19-21.
The Court finds that Plaintiffs have articulated a damages model that is capable of calculating damages on a class-wide basis. First, “Plaintiffs' sole theory of liability” – misrepresentations regarding the Kemper Plant – “giving rise to their sole alleged harm, artificial stock inflation, is straightforward.” Beaver Cty., 2016 WL 4098741, at *11. Second, Professor Feinstein's proposed damages model has been accepted in securities fraud class actions in cases like this one, where Plaintiffs seek out-of-pocket damages to recover artificial inflation caused by Defendants' misrepresentations. See, e.g., Wilson, 2018 WL 3913115, at *16 (finding Professor Feinstein's identical methodology satisfied Rule 23(b)(3)); Luna v. Marvell Tech. Grp., Ltd., No. C. 15-05447 WHA, 2017 WL 4865559, at *6 (N.D. Cal. Oct. 27, 2017) (“The use of an event study to isolate damages stemming from a particular cause is not unique to this action. It is a feature of virtually every securities action, which must account for stock fluctuations unrelated to the particular theory of liability asserted in the case.”); W. Va. Pipe Trades Health & Welfare Fund v. Medtronic, Inc., 325 F.R.D. 280, 290 (D. Minn. 2018) (upholding Professor Feinstein's proposed event study methodology for classwide damages); In re Silver Wheaton Corp. Sec. Litig., No. 2:15-cv-05146-CAS (JEMx), 2017 WL 2039171, at *14-*15 (C.D. Cal. May 11, 2017) (same).
Defendants did not move to exclude Professor Feinstein's damages opinion. Rather, Defendants and Professor Gompers criticize Professor Feinstein's model as “boilerplate” and argue that Professor Feinstein has proposed this same model in dozens of other securities cases. Defendants' Class Certification Opposition at 21; Day 1 Tr. at 171:2-12. Defendants also argue that Professor Feinstein's model is too vague to demonstrate that damages can be calculated on a class-wide basis because it does not specify the valuation tools that will be used to implement the model. Id. They point to the Northern District of Ohio's opinion in Freddie Mac, 2018 WL 3861840, which excluded Professor Feinstein's damages opinion, and urge the Court to find that Professor Feinstein has failed to articulate a class-wide damages model here.
The Court declines to do so. First, although Professor Feinstein has not yet specified which valuation tools – an input into the damages model – he will ultimately use, such specification is not required at this stage. See Wilson, 2018 WL 3913115, at *17 (rejecting defendants' criticism that “Feinstein does not specify which ‘valuation tools’ he will use, how he will ‘separate the effect’ ” of corrective disclosures and materializations of the risk, or “ ‘how he will calculate inflation’ ”). Nevertheless, Professor Feinstein testified in this case how, based on developments on the merits, he may utilize valuation tools (such as revenue, book value, cash flow and discounted cash flow models based on earnings or return attribution analysis) to calculate inflation in his damages model. Day 1 Tr. at 130:10-23. Because the decision of which, if any, of those tools will be necessary to measure damages in this case depends on development of the fact record on the merits, it would be inappropriate for Professor Feinstein to conclusively state which he would use at this stage of the litigation. Wilson, 2018 WL 3913115, at *17.36 It is sufficient for class certification that Professor Feinstein has specified a damages model that can be used to establish damages using a common methodology for all class members, even though certain of the inputs to that model are not yet ascertainable. See Thorpe, 2016 WL 4006661, at *16 (“ ‘nothing in Comcast requires an expert to perform his analyses at the class certification stage’ ”).
Defendants further argue that Professor Feinstein's model does not match Plaintiffs' theory of liability, which Defendants argue is a “materialization of the risk theory” that requires calculating inflation based on a variable probability of risk that the Kemper Plant would not be completed on time. See [Doc. 106 at 22]. Defendants argue that there is a possibility of a windfall recovery, in violation of Comcast, if Plaintiffs recover damages caused by the announcement that the Kemper Plant would not meet the May 204 COD, if, during the Class Period, it was not a certainty that the Kemper Plant would miss the May 2014 COD. Id. at 23-24.
Plaintiffs argue that the Defendants misconstrue Plaintiffs' theory of liability. Although the phrase “materialization of the risks concealed” is present in the Complaint, it appears in the “Loss Causation” section of the Complaint and describes the manner in which corrective information was revealed to the market. Plaintiffs argue that the phrase does not describe Plaintiffs' theory of liability. Rather, Plaintiffs' theory of liability is that, from the first day of the Class Period, there was a 100% likelihood that the Kemper Plant would not be completed on time. See Defendants' Class Certification at 13-14. According to Plaintiffs, because it was “impossible” that the Company would achieve the stated completion date as of the start of the Class Period, it was a fact that the Kemper Plant would not be completed by the May 2014 COD. Id.; see also, e.g., ¶¶6, 68, 78, 115(e), 124(e)-(f), 131(f)-(g), 144(f), and 159 (e)-(f) (alleging that achieving the May 2014 COD was “impossible” throughout the Class Period). Plaintiffs similarly allege that Defendants' statements that the Kemper Plant was “on track,” “on schedule,” “70 percent” and “75 percent” complete, and that certain construction milestones had been achieved were objectively false when made – at the time Defendants represented as much, it was a fact that construction was not “on track,” “on schedule,” “70 percent” or “75 percent” complete, and construction milestones had not been achieved. These facts were, according to Plaintiffs, known to Defendants at the time they uttered the alleged misrepresentations and omissions.
Accordingly, Professor Feinstein testified at his deposition that his “understanding of the case is that plaintiffs alleged not that the probability of missing the deadline kept going up, but that the likelihood of meeting the deadline was virtually nil from the start, but that was concealed.” Day 1 Tr. at 105:5-106:3, 157:23-158:16; see also id. at 105:17-25 (“[F]rom the outset of the class period the allegation is that the company knew they were going to miss that deadline, the commercial operation date deadline for the plant. So it wasn't that the probability of missing the deadline kept growing over time. That's not the allegation. The allegation is that it was known to the company that they would necessarily miss the deadline, and they concealed that fact from the public.”). Consistent with Plaintiffs' theory of liability, Plaintiffs contend that Professor Feinstein's damages model will calculate damages based on the difference between the observed stock price and the price the stock would have traded at had the market known that achieving the May 2014 COD was “impossible” from the start. See, e.g., Day 1 Tr. at 117:2-5, 121:7-10, 19-21; [Doc. 113-2, ¶¶234-238].37
Professor Gompers' criticisms on this point, which he concedes are based on an understanding of Plaintiffs' legal theory provided by defense counsel, are inapposite. See Day 1 Tr. at 171:13-175:22, 240:24-241:1 (Professor Gompers did not read Judge Cohen's opinion to determine how the court had characterized Plaintiffs' allegations), 241:6-16 (Professor Gompers conceded that Plaintiffs' theory of liability was “related to [him]”). In fact, Professor Gompers highlighted his misunderstanding of Plaintiffs' allegations at the evidentiary hearing when he testified that it was “not [his] understanding” that Plaintiffs allege there was a 100% chance the Kemper Plant would not be completed by the May 2014 COD. Day 1 Tr. at 242:7-14. Professor Gompers also admitted in his deposition that if Plaintiffs' allegations were that “the true risk of the Kemper Plant not being completed was a hundred percent and some interim partial disclosure occurs that relates to a possibility that the schedule might not be met,” then “that's fraudulently saying something is not going to happen․ when you know for sure it's going to happen. So that's not a materialization of a risk.” Day 1 Tr. at 243:6-18; Plaintiffs' Ex. 6, Gompers Tr. at 167:13-168:2
Defendants' citation to a handful of cases finding that plaintiffs' damages model did not account for “materialization of the risk” issues highlight why this case is different. In In re BP p.l.c. Sec. Litig., No. 4:10-md-2185, 2014 WL 2112823 (S.D. Tex. May 20, 2014) (“BP II”), aff'd sub nom. Ludlow v. BP, P.L.C., 800 F.3d 674 (5th Cir. 2015), for example, the plaintiffs sought to certify two separate classes of investors – those who purchased pre-oil spill and those who purchased post-oil spill – alleging two separate theories of liability and damages models. Id. at *2-*4. The pre-spill subclass alleged a “materialization of the risk” theory of liability; namely, that they had purchased in reliance on the defendants' representations that BP maintained its oil rigs safely and in compliance with all applicable regulations, thus concealing the true risk of a future oil spill. The district court declined to certify the pre-spill subclass because plaintiffs' “materialization of the risk” theory was inconsistent with their proffered damages model, which could not account for the fact that the underlying risk of a future oil spill changed throughout the class period. Id. at *2-*4, *12.
Plaintiffs' theory of liability is straightforward, and their class-wide damages model is appropriately articulated at this stage. Moreover, Plaintiffs do not allege that the underlying risks vis-à-vis the Kemper Plant changed over time. The Court finds that Plaintiffs' proffered damages model satisfies Rule 23(b)(3).
2. Superiority
Before certifying a class under Rule 23(b)(3), the Court must find that a class action is “superior to other available methods for fairly and efficiently adjudicating the controversy.” Rule 23(b)(3). To make this decision, the Court considers:
(A) the class members' interests in individually controlling the prosecution or defense of separate actions; (B) the extent and nature of any litigation concerning the controversy already begun by or against class members; (C) the desirability or undesirability of concentrating the litigation of the claims in the particular forum; and (D) the likely difficulties in managing a class action.
Id.
Defendants do not contest that a class action is the superior method for adjudicating this case, and the Court finds this requirement satisfied. Absent class members have little interest in individually controlling separate actions, as the amount of individual damages is likely to be relatively small. The Court is not aware of any securities fraud litigation concerning this controversy that has already begun by or against absent class members. Concentrating litigation of the claims in this forum is desirable, as Southern Company is headquartered in this District and many of the acts alleged in the Complaint occurred in this District. Finally, any difficulties in managing a class action do not outweigh the benefits of certifying a class in this case. Accordingly, the Court finds that adjudicating this case as a class action is the superior method.
V. CONCLUSION
For the foregoing reasons, the Court finds that Plaintiffs have established the requirements for class certification under Rule 23(a) and 23(b)(3). Consequently, Plaintiffs' motion for class certification and appointment of Class Representatives and Class Counsel is GRANTED. The Court certifies the following Class:
All persons who purchased or otherwise acquired The Southern Company common stock between April 25, 2012 and October 30, 2013, inclusive (the “Class Period”), and were damaged thereby.38
Roofers Local No. 149 and Monroe County are appointed as Class Representatives, and Robbins Geller is appointed as Class Counsel.
IT IS SO ORDERED, this 22nd day of August, 2019.
FOOTNOTES
1. Unless noted otherwise, all “¶_” and “¶¶_” references are to the Complaint.
2. Citations are omitted and emphasis is added, unless otherwise noted.
3. Excluded from the Class are Defendants, the officers and directors of Southern Company, at all relevant times, 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. at 1 n.1.
4. For class certification purposes, the Supreme Court has found that falsity, materiality, and loss causation are issues common to a class because “failure of proof” on any of these elements “would end the case” for all putative class members. Amgen, 568 U.S. at 468, 475, 133 S.Ct. 1184 (“this Court has held that loss causation and the falsity or misleading nature of the defendant's alleged statements or omissions are common questions that need not be adjudicated before a class is certified”; “the element of materiality would end the case for one and for all”); Halliburton I, 563 U.S. at 813, 131 S.Ct. 2179 (loss causation is a common question and plaintiffs are not required to “show loss causation as a condition of obtaining class certification”).
5. Major institutions are defined as “firms or individuals that exercise investment discretion over the assets of others in excess of $100 million.” [Doc. 77-2., ¶62].
6. Not only do Defendants fail to rebut or even address any of the “traditional indicia of efficiency,” but their expert – Professor Gompers – conceded at the evidentiary hearing, “I think that most of the time for [firms trading on the NYSE] it's likely that they're efficient.” [Doc. 135 at 180:1-2] (Amended Transcript of the May 21, 2019 Proceedings Before The Honorable William M. Ray, II (“Day 1 Tr.”)).
7. “As stated by the SEC:․ Proposed Form S-3 recognizes the applicability of the efficient market theory to the registration statement framework with respect to those registrants which usually provide high quality corporate reports, including Exchange Act reports, and whose corporate information is broadly disseminated, because such companies are widely followed by professional analysts and investors in the market place․ Because of the foregoing observations made by the SEC, the existence of Form S-3 status is an important factor weighing in favor of a finding that a market is efficient.” Cammer, 711 F. Supp. at 1284-85.
8. Accordingly, district courts around the country routinely find market efficiency regardless of the fifth Cammer factor. See, e.g., NetBank, 259 F.R.D. at 669, 674-75 (“Proof of every single factor is not always necessary․ Defendants' arguments on the [fifth] factor are insufficient to overcome a prima facie showing of market efficiency at the class certification stage.”); In re Scientific-Atlanta, Inc. Sec. Litig., 571 F. Supp. 2d 1315, 1339 (N.D. Ga. 2007) (presumption not rebutted where defendants challenged only the fifth Cammer factor); City of Cape Coral Mun. Firefighters' Ret. Plan v. Emergent Biosolutions, Inc., 322 F. Supp. 3d 676, 689 (D. Md. 2018) (explaining that “[e]ven if the Court were to find [plaintiffs' expert's] opinion regarding the fifth, empirical Cammer factor inadmissible, the Court would still find” plaintiffs established market efficiency); Angley v. UTI Worldwide Inc., 311 F. Supp. 3d 1117, 1121 (C.D. Cal. 2018) (“Because there is no evidence disputing the first four Cammer factors and the Krogman factors weigh in favor of market efficiency, the Court finds Plaintiff has met its burden of showing market efficiency.”); In re NII Holdings, Inc. Sec. Litig., 311 F.R.D. 401, 412 (E.D. Va. 2015) (holding fifth Cammer factor is not “dispositive of the market efficiency inquiry”); Wilson v. LSB Indus., Inc., No. 15 Civ. 7614 (RA) (GWG), 2018 WL 3913115, at *9 (S.D.N.Y. Aug. 13, 2018) (“no one factor is dispositive”); W. Palm Beach Police Pension Fund v. DFC Glob. Corp., No. 13-6731, 2016 WL 4138613, at *12 (E.D. Pa. Aug. 4, 2016) (“Courts have rejected the idea that the fifth Cammer factor is necessary to establish market efficiency.”); In re Montage Tech. Grp. Ltd. Sec. Litig., No. 14-cv-00722-SI, 2016 WL 1598666, at *8 (N.D. Cal. Apr. 21, 2016) (“absence of any one Cammer factor is not determinative”); Första AP-Fonden v. St. Jude Med., Inc., 312 F.R.D. 511, 520 (D. Minn. 2015) (same); Beaver Cty. Emps.' Ret. Fund v. Tile Shop Holdings, Inc., No. 14-786 ADM/TNL, 2016 WL 4098741, at *11 (D. Minn. July 28, 2016) (same); Smilovits v. First Solar, Inc., 295 F.R.D. 423, 437 (D. Ariz. 2013) (same).
9. The finance literature “notes that [unexpected, value-relevant] information more frequently arrives on earnings announcement dates than on ordinary dates.” [Doc. 77-2, ¶101]. “Analysts, investors, senior executives, and boards of directors consider earnings the single most important item in the financial reports issued by publicly held firms.” “Earnings Management to Exceed Thresholds,” by Francois Degeorge, Jayendu Patel, and Richard Zeckhauser, Journal of Business, 1999, p.1.
10. Defendants' reliance on Ohio Pub. Emps. Ret. Sys. v. Fed. Home Loan Mortg. Corp., No. 4:08CV0160, 2018 WL 3861840 (N.D. Ohio Aug. 14, 2018) (“Freddie Mac”), to critique Professor Feinstein's market efficiency opinion misses the mark. In Freddie Mac, the court was critical of Professor Feinstein's use of only one event date. Id. at *3-*4. Here, however, Professor Feinstein used seven earnings announcement dates by Southern Company.
11. Of the two event dates that were not followed by statistically significant price reactions using one-day windows, one of the dates – January 30, 2013 – was deemed by Professor Feinstein to be a “mixed” news announcement, “meaning that a statistically significant price reaction would not be expected. Excluding that date, five of six announcements that would be expected to cause a statistically significant price reaction were, in fact, followed by a statistically significant price reaction using one-day windows.” [Doc. 113-2, ¶91 n.76] (Rebuttal Report of Professor Steven P. Feinstein, Ph.D., CFA, dated Mar. 29, 2019 (“Feinstein Rebuttal”)). Professor Gompers does not challenge Professor Feinstein's determination that the news announced on January 30, 2013 was “mixed” such that a significant price reaction would not be expected.
12. There is no requirement that every event date tested be followed by a statistically significant price reaction in order to conclude that the market for a stock was efficient. See, e.g., Regions, 762 F.3d at 1257 (rejecting argument that plaintiff must “prove a set number of unexpected disclosures resulting in an immediate price impact” in order to establish market efficiency); Thorpe, 2016 WL 4006661, at *13 (expert's finding that “stock price ‘reacted strongly’ ” following five of 10 earnings announcements supported market efficiency); In re Petrobras Sec. Litig., 312 F.R.D. 354, 371 (S.D.N.Y. 2016) (“sid[ing] with Feinstein” over Gompers, finding that “not every event will move a market and that the impact of an event depends on various factors,” and noting that “[t]he Supreme Court has rejected Gompers' absolutist view of market efficiency by making clear that ‘market efficiency is a matter of degree’ ”), aff'd in part and vacated in part on other grounds, 862 F.3d 250 (2d Cir. 2017).
13. At the evidentiary hearing, Professor Gompers testified that “no financial economist would run a Fisher's Exact Test to assess market efficiency.” Day 1 Tr. at 227:22-24. But when confronted with his 2013 report in Deutsche Bank, in which Professor Gompers ran a Fisher's Exact Test to argue that the number of significant price reactions proved market inefficiency, Professor Gompers admitted that he too has employed Fisher's Exact Tests to assess market efficiency. See IBEW Local 90 Pension Fund v. Deutsche Bank AG, No. 11-cv-4209 (KBF), ECF No. 62 (S.D.N.Y. Aug. 29, 2013).
14. The October 30, 2013, earnings announcement elicited significant price reactions using both a one- and two-day event window. [Doc. 77-2, ¶¶163-164]. Thus, the Court addresses only Professor Feinstein's analysis of a two-day event window following the July 31, 2013, earnings announcement.
15. Although Southern Company issued a press release with bottom-line earnings information before the market opened on July 31, 2013, Professor Feinstein explained that the analyst reports that were issued before the 1:00 p.m. Eastern earnings call “indicated that analysts were awaiting further details concerning the additional Kemper Plant charges.” [Doc. 113-2, ¶86]. For example, among the few pre-call analyst reports was a report from Citi Research, which stated: “ ‘We look for the earnings call tomorrow for further clarity and details.’ ” Id. Similarly, Atlantic Equities' pre-call report on July 31, 2013, noted the additional Kemper Plant charge and indicated that it awaited information concerning the effect on financial guidance, which would be provided on the earnings call: “We note that the company typically discussed FY guidance on the call but does not traditionally adjust guidance until the third quarter.” Id. UBS's pre-call report on July 31, 2013, also indicated that it awaited “[d]iscussion” of the “Kemper IGCC construction and cost update” on the “[c]all at 1pm ET.” Id.
16. Professor Feinstein explained that because the market had only 90 minutes to process this “blockbuster negative news[,] [i]t makes sense that some of [the price reaction] would spill over to August 1st.” Day 1 Tr. at 91:22-92:1. In response, Defendants cited the Company's disclosure months earlier that it might need to raise $0 to $300 million in equity to argue that the July 31st revelation that the Company would in fact raise $700 million due to previously concealed delays and cost overruns at the Kemper Plant was not “new” news. But the Company's July 31st revelation of the certainty of $700 million in equity dilution as a direct result of the Kemper Plant problems was markedly different from its earlier warning that the Company may need to issue equity in the amount of $0 to $300 million. In fact, analysts pointed to the Company's earlier statement that Southern Company may need to raise $0 to $300 million to underscore their shock at the dramatically different July 31st announcement that Southern Company required $700 million in additional equity. See, e.g., [Doc. 113-2, ¶87]. Further, Professor Gompers testified that he did not analyze whether new, value-relevant information was released on the July 31, 2013 earnings call shortly before the market closed or “ ‘tabulate’ ” “ ‘whether analysts’ reports that were issued” after the market closed for those event dates “ ‘reported on and discussed new information that they learned during those calls.’ ” [Doc. 114-1 at 13-14] (Plaintiffs' Memorandum of Law in Support of Motion to Exclude Report and Opinions of Dr. Paul A. Gompers). Therefore, the Court gives little weight to Professor Gompers' opinion that no new information was discussed during the July 31, 2013, earnings call.
17. One variable in an event study model is the sector index, which “control[s] for the effects of market-wide and industry news on the company's stock price, so that one can isolate the effect of company-specific news.” [Doc. 113-2, ¶99].
18. Professor Feinstein conducted a Fisher's Exact Test to test the probability of observing three of seven earnings announcement eliciting statistically significant one-day stock price reactions. Professor Feinstein concluded that the probability of observing Professor Gompers' results (i.e., three of seven event dates eliciting statistically significant price reactions) in an inefficient market would be only 0.502%. [Doc. 113-2, ¶95]. Thus, at a 99.498% confidence level, Professor Gompers' event seemingly study rejects the hypothesis that Southern Company stock did not trade in an efficient market. Id.
19. Although event studies are not required in this Circuit, Professor Gompers opined that “the only evidence a financial economist would look to in terms of offering․ his conclusion is the direct test, an event study.” Day 1 Tr. at 198:23-199:1. Professor Gompers readily testified that he was unaware that event studies are not required in this Circuit (id. at 212:11-14) and that it does not matter to him in offering opinion testimony in this case “how Courts state that market efficiency can be established.” Id. at 217:2-4. But though “[i]t may be that many financial economists, including Dr. Gompers,” believe that event studies are required, such a position “ ‘describe[s] a different conception of an efficient market than is used by the law.’ ” Willis, 2017 WL 1074048, at *4.
20. SEC Regulation S-K provides publicly traded companies wide latitude in selecting a sector index, inviting companies to compare themselves against a “published industry or line-of-business index”; “[p]eer issuer(s) selected in good faith”; or “[i]ssuer(s) with similar market capitalizations, but only if the registrant does not use a published industry or line-of-business index and does not believe it can reasonably identify a peer group.” 17 C.F.R. § 229.201. Moreover, to the extent a company has any difficulties in identifying an appropriate industry sector index, SEC Regulation S-K explicitly invites companies to describe those difficulties or deficiencies or to construct a custom index out of “peers issuers selected in good faith.” Id. Although SEC Regulation S-K permitted Southern Company to present an index comprised of “[p]eer issuer(s) selected in good faith,” Southern Company chose not to do so. Id. Rather, Southern Company presented a “published industry or line-of-business index” – the S&P EUI.
21. Standard & Poor's excluded from the S&P EUI companies deemed to be “multi-utility” companies, which are companies that provide other, non-electric utility services, such as water and natural gas. Id.
22. Professor Gompers relies on “Econometrics: Legal, Practical, and Technical Issues” by Lawrence Wu (ABA Antitrust Section, 2014) (“Wu, Econometrics”) to explain the meaning of a higher R2. [Doc. 106-2, ¶57 & nn.134-135]. But Professor Gompers omits Wu's conclusion that “[a] model with a high R-squared statistic is not necessarily better than a different model with a lower R-squared statistic, and one should not place too great of importance on finding a model with a high R-squared statistic.” Wu, Econometrics at 100. Professor Gompers also relies on “Introduction to Econometrics,” by James H. Stock and Mark W. Watson, 2nd ed. (Boston: Pearson/Addison-Wesley, 2007) (“Introduction to Econometrics”) to explain the concept of fit. [Doc. 106-2, ¶57 & nn.131-133, 135]. But Professor Gompers omits the article's conclusion that “heavy reliance on the [R value, or fit] can be a trap. [Seeking to] ‘maximize the [R value]’ is rarely the answer to any economically or statistically meaningful question.” Introduction to Econometrics at 202.
23. In his first report, Professor Gompers indicated that that the “Custom Peer Group” was an index used by Southern Company. [Doc. 106-2, ¶15]. At his deposition, Professor Gompers surmised that although Southern Company's proxy statement “doesn't mention [a Custom Peer Group] index, it's [his] belief that [Southern Company] had to have created [one].” Plaintiffs' Ex. 6, Gompers Tr. at 175:13-18. On rebuttal, Professor Gompers abandoned his claim that the “Custom Peer Group” was an index, stating instead that “[w]hether the Custom Peer Group is referred to as an ‘index’ in Southern's SEC filings is irrelevant.” [Doc. 117-2] (Declaration of Paul A. Gompers, dated Apr. 12, 2019 (“Gompers Rebuttal”)), ¶34. At the evidentiary hearing, Professor Gompers ultimately conceded that Southern Company did not use the Custom Peer Group as an index – “[t]hey did not create a composite number from those companies.” Day 1 Tr. at 236:4-5; id. at 236:10-11 (“I'm unaware of them calculating a single return series from the custom peer group.”).
24. “Event Studies in Economics and Finance” by Craig A. MacKinlay, Journal of Economic Literature, Vol. 35, No. 1, 1997, pp. 14-15. MacKinlay makes clear that examination of stock price effects that occur on the second trading day is perfectly appropriate in event study analysis.
25. “Materiality and Magnitude: Event Studies in the Courtroom,” by David I. Tabak and Frederick C. Dunbar, Litigation Services Handbook, The Role of the Financial Expert, 3d ed., 2001, p. 194; see also Feinstein Rebuttal, ¶¶53-57, 63, 66-68 (citing “Does M&A Pay? A Survey of Evidence for the Decision-Maker,” by Robert F. Bruner, Journal of Applied Finance, Spring/Summer 2002; “The Intraday Speed Of Adjustment Of Stock Prices To Earnings And Dividend Announcements,” by James Patell and Mark Wolfson, Journal of Financial Economics, Vol. 13, pp. 223-252, 1984, p. 235; “Intradaily Price-Volume Adjustments of NYSE Stocks to Unexpected Earnings,” by C. Woodruff and A. Senchack, The Journal of Finance, 1988, pp. 487-490; “Price Discovery on the NYSE and the NASDAQ: The Case of Overnight and Daytime News Releases,” by Jason Greene and Susan Watts, Financial Management, Vol. 25, No. 1, Spring 1996, pp. 19-42, p. 37; “Announcement Effects Of New Equity Issues And The Use Of Intraday Price Data,” by Michael Barclay and Robert Litzenberger, Journal of Financial Economics, Vol. 21, 1998, pp. 71-99, p. 92; “Market Efficiency In Real Time,” by Jeffrey Busse and T. Clifton Green, Journal of Financial Economics, Vol. 65, 2002, pp. 415-437, p. 416).
26. See Day 1 Tr. at 141:21-25 (explaining that 90 minutes is “a short amount of time to analyze the blockbuster news that the company was going to have to issue $700 million in․ new equity, diluting its equity in order to respond to overruns at Kemper”); id. at 144:20-25 (“[T]he conference call is where they disclosed the dilution and the new $700 million equity issue that the company would have to execute in order to raise the money to finish the plant. So that was blockbuster news at the end of the trading day on July 31st.”).
27. With respect to Defendants' argument that Professor Feinstein was inconsistent because he did not examine the second day for certain other dates, Professor Feinstein explained that an event study analysis is an investigation to confirm that a company's stock price reacts to news. Where there was a one-day significant price reaction, Professor Feinstein explained that there was no need to investigate the second day because the event study confirmed that the market responded to news. Day 1 Tr. at 146:13-148:25. Of course, the trier of fact will ultimately get the final say as to whether Professor Feinstein's approach was reasonable.
28. Because Defendants' allegedly false statements and omissions regarding the construction and status of the Kemper Plant were not unexpected, they would not be anticipated to cause a statistically significant stock price increase. [Doc 113-2, ¶¶95-99]. As the Eleventh Circuit held in Regions, “[w]hen a company releases expected information, truthful or otherwise, the efficient market hypothesis underlying Basic predicts that the disclosure will cause no significant change in the price.” 762 F.3d at 1256. Consequently, for purposes of price impact, Defendants focus on the stock price reaction following the alleged corrective disclosures.
29. Alon Brav was one of Professor Gompers' Ph.D. students at The University of Chicago, and Professor Gompers has written articles with Dr. Brav. Day 1 Tr. at 249:21-22.
30. During the evidentiary hearing, Professor Feinstein provided an analogy to this situation: “[I]f you want to ascertain whether or not a particular lake has fish in it, you might want to throw a line in and see if you can catch a fish. If you catch a fish, that's proof that the lake has fish. If you don't catch a fish, that hasn't proved anything. You haven't proved that there's no fish in the lake. Might prove that you're a bad fisherman and that you threw the line into the wrong place.” Day 1 Tr. at 96:22-97:3.
31. During the evidentiary hearing, counsel for Defendants noted that the defendants in Rooney sought review of the class certification order by the Fifth Circuit. Subsequently, however, the parties in Rooney informed the court that they settled the case and “jointly moved the United States Court of Appeals for the Fifth Circuit to hold in abeyance Defendants' pending appeal” of the class certification order. In re EZCORP, Inc. Sec. Litig., No. 1:15-cv-00608-SS, ECF No. 134 (W.D. Tex. June 5, 2019).
32. Defendants cite Erica P. John Fund, Inc. v. Halliburton Co. (Halliburton III) as support for their argument that a non-statistically significant decline following a corrective disclosure proves an absence of price impact. 309 F.R.D. 251, 280 (N.D. Tex. 2015). However, the court there based its ruling both on the absence of a statistically significant price reaction and that the relevant information had been previously disclosed. Defendants also cite In re Intuitive Surgical Securities Litigation, 2016 WL 7425926, *15-*16 (N.D. Cal. 2016), in which plaintiffs conceded that a non-statistically significant decline would prove no price impact. Thus, the court did not resolve that question but, rather, cited to Halliburton III without analysis. Defendants' other case, In re Finisar Corporation Securities Litigation, 2017 WL 6026244 (N.D. Cal. 2017), did not analyze the question of whether a non-significant decline proves an absence of price impact.
33. Not only did the price of Southern Company's stock decline on an absolute dollar basis following each of the five alleged corrective disclosures, but Professor Feinstein demonstrated that there was a negative residual return following each of those disclosures. See [Doc. 113-2 p. 110]. This establishes that Southern Company's stock price experienced a decline after factoring out market and sector effects. In other words, Southern Company's residual stock price decline was larger than the return predicted by the market and its sector peers. Notably, even Professor Gompers' event study found negative residual returns following each of the five corrective disclosures. See [Doc. 106-2, p. 141].
34. During the evidentiary hearing, Defendants argued that Halliburton II “says we can rely on event studies too” in order to prove an absence of price impact. Day 2 Tr. at 91:23-24. First, Halliburton II did not hold that event study analysis finding a non-significant price decline following a corrective disclosure proves the absence of price impact. Second, Plaintiffs do not argue that Defendants may not submit an event study for purposes of demonstrating no price impact. Rather, Plaintiffs argue that even Defendants' event study establishes that Southern Company's stock price experienced a negative residual price decline following every single one of the alleged corrective disclosures. See [Doc. 106-2, p.141]. That fact, combined with the fact that a non-significant decline does not allow one to rule out the possibility that the stock price decline that did occur was unrelated to the alleged misrepresentations, leads the Court to conclude that Defendants have failed to establish that there was “no price impact whatsoever.” Emergent Biosolutions, 322 F. Supp. 3d at 687.
35. Defendants cite no binding authority, and the Court is aware of none, requiring Plaintiffs to produce a damages model in order to satisfy predominance. See, e.g., Scientific-Atlanta, 571 F. Supp. 2d at 1342-43 (certifying class where expert did not even provide a damages model, stating only “that he ‘believe[s] there are tried and tested methods and procedures by which damages of class members can be computed on a formulaic class-wide basis, with relatively little analysis’ ” because “aggregate damages models have been employed in numerous other federal securities cases”); In re Barrick Gold Sec. Litig., 314 F.R.D. 91, 105-06 (S.D.N.Y. 2016) (rejecting Comcast challenge and granting class certification where plaintiffs did not proffer a damages model but stated they would rely on an out-of-pocket theory of damages).
36. Defendants suggest that Professor Feinstein's damages opinion is inadequate because he has not yet performed his damages analysis. At the evidentiary hearing, for example, Defendants suggested that Professor Feinstein's damages opinion is inadequate because he could not identify, at this point, the amount of inflation that was in the stock on a particular day during the Class Period. See Day 1 Tr. at 112:5-7, 129:9-12. But Plaintiffs are not required at the class certification stage to perform a damages analysis. Professor Feinstein testified that he could prepare a schedule of inflation, and that for any particular investor, the claims administrator would be able to use that schedule, and the dates the class member purchased and sold their Southern Company shares, and determine the difference between the amount of inflation in the stock on those particular days. That difference would allow the individual shareholder's damages to be calculated on an individual basis, though the same damages formula is applicable to all class members. Id. at 160:15-161:4. Nothing more is required at this stage.
37. Although Plaintiffs make clear that this case alleges a 100% likelihood that the Kemper Plant would not be completed by the May 2014 COD and that construction was not “on target” when Defendants represented as much, Professor Feinstein nonetheless testified that even if one was required to address a varying level of underlying risk for purposes of damages, that is “not even necessarily all that unusual․ [T]here are a lot of cases like that where the probability of an adverse event is changing over time. One uses the valuation tools to evaluate at all points in time what would the stock have been worth if what was known to the company at that point in time was revealed to the marketplace.” Day 1 Tr. at 106:9-14; see also id. at 126:12-16 (“I can tell that you the damage model accommodates allegations that the company knew with certainty that they would miss the deadline just as it accommodates allegations that there was an elevated level of risk and that that elevated level of risk may have changed.”).
38. Excluded from the Class are Defendants, the officers and directors of Southern Company, at all relevant times, 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.
WILLIAM M. RAY, II, United States District Judge
A free source of state and federal court opinions, state laws, and the United States Code. For more information about the legal concepts addressed by these cases and statutes visit FindLaw's Learn About the Law.
Docket No: CIVIL ACTION NO: 1:17-cv-00241-WMR
Decided: August 22, 2019
Court: United States District Court, N.D. Georgia, Atlanta Division.
Search our directory by legal issue
Enter information in one or both fields (Required)
Harness the power of our directory with your own profile. Select the button below to sign up.
Learn more about FindLaw’s newsletters, including our terms of use and privacy policy.
Get help with your legal needs
FindLaw’s Learn About the Law features thousands of informational articles to help you understand your options. And if you’re ready to hire an attorney, find one in your area who can help.
Search our directory by legal issue
Enter information in one or both fields (Required)