Summary
finding that "event study methodology met Daubert's exacting standards"
Summary of this case from Ploss v. Kraft Foods Grp.Opinion
Master File No. 12 C 2450
03-06-2015
In re GROUPON, INC. SECURITIES LITIGATION, This Document Relates To: ALL ACTIONS.
MEMORANDUM OPINION AND ORDER
Defendants have moved to exclude the opinions and testimony of Steven P. Feinstein. For the reasons set forth below, the Court denies the motion.
I. BACKGROUND
Lead Plaintiff asserts Securities Act claims, alleging that Defendants made false statements in Groupon's IPO registration statement and Exchange Act claims, based on Groupon's reported financial results from the fourth quarter of 2011, as well as oral statements allegedly made by some of the individual defendants in February and March 2012. Plaintiff seeks to represent: (a) all persons or entities who purchased or acquired the common stock of Groupon pursuant and/or traceable to the IPO between November 4, 2011 through and including March 30, 2012 (the Securities Act Class); and (b) a subclass consisting of all persons who purchased or otherwise acquired the common stock of Groupon between February 9, 2012 through and including March 30, 2012 (the Exchange Act Subclass).
For purposes of the Exchange Act Subclass only, Plaintiff relies on the opinions and testimony of Dr. Steven P. Feinstein to establish that Groupon traded in an efficient market. (Dkt. 188 at 19-23). Defendants have moved to exclude Dr. Feinstein's opinions and testimony, arguing that they fail to meet the legally required standards for expert testimony under Fed. R. Evid. 702 and Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579 (1993). (Dkt. 235). Defendants contend that their expert, Dr. Paul A. Gompers, has demonstrated that Dr. Feinstein's analysis is unsound and unreliable. (Id. at 5-13).
On December 4, 2013, Dr. Feinstein submitted an expert report, opining that Groupon stock traded in an efficient market over the course of the Subclass period. (Dkt. 189, Ex. 1 at ¶ 19). On March 5, 2014, Dr. Gompers submitted an expert report, concluding that Dr. Feinstein's opinion was based on an incomplete and flawed analysis, rendering Dr. Feinstein's opinion unreliable. (Dkt. 212, Ex. A at ¶ 7). On September 11, 2014, this Court conducted a Daubert evidentiary hearing at which both Drs. Feinstein and Gompers testified and were cross-examined. (Dkt. 253; see Dkt. 273 Ex. 1 (Hr'g Tr., Sept. 11, 2014)).
On September 23, 2014, the District Court certified both the Securities Act class and the Exchange Act Subclass pursuant to Federal Rule of Civil Procedure 23(b)(3). (Dkt. 254). On October 1, 2014, Defendants filed a Rule 23(f) petition requesting the Seventh Circuit to vacate and remand the class certification order so that the Daubert analysis could be concluded. (See Dkt. 255 at ¶ 8). On October 7, 2014, this Court granted Defendants' motion to stay the post-Daubert hearing briefing. (Dkt. 260).
On November 24, 2014, the Seventh Circuit held Defendants' Rule 23(f) petition in abeyance, "so that the district court has an opportunity to address Defendants' motion to exclude the opinions and testimony of Dr. Feinstein." (See Dkt. 261 Ex. A). On December 5, 2014, the District Court granted Defendants' motion to permit resumption of the post-Daubert briefing, and this Court lifted the stay. (Dkt. 271, 272). Defendants filed their post-Daubert hearing supplemental brief on December 19, 2014. (Dkt. 273). Plaintiff filed his opposition on January 8, 2015 (Dkt. 275), and on January 20, 2015, Defendants filed a reply (Dkt. 279).
II. DISCUSSION
A. Legal Standard for Daubert Motions
"While expert testimony can be useful to both courts and juries in securities cases, an expert must, in fact, be an expert; and, his or her opinions must demonstrate such expertise." IBEW Local 90 Pension Fund v. Deutsche Bank AG, No. 11 C 4209, 2013 WL 5815472, at *12 (S.D.N.Y. Oct. 29, 2013) (footnote omitted). Federal Rule of Evidence 702 provides that
a witness who is qualified as an expert by knowledge, skill, experience, training, or education may testify in the form of an opinion or otherwise if:
(a) the expert's scientific, technical, or other specialized knowledge will help the trier of fact to understand the evidence or to determine a fact in issue;
(b) the testimony is based on sufficient facts or data;
(c) the testimony is the product of reliable principles and methods; and
(d) the expert has reliably applied the principles and methods to the facts of the case.The Supreme Court has tasked the district court with the responsibility for "ensuring that an expert's testimony both rests on a reliable foundation and is relevant to the task at hand." Daubert, 509 U.S. at 597. This gatekeeping role mandates "that a court should not simply accept a proffered expert—even if he has been accepted as such before." Deutsche Bank, 2013 WL 5815472, at *12. Instead, the court must conduct an inquiry into whether an expert is an expert, and exercise some degree of "regulation of the subjects and theories about which an expert may testify." Daubert, 509 U.S. at 589; see Deutsche Bank, 2013 WL 5815472, at *12.
The Daubert inquiry requires a court to determine: "(1) whether the proposed expert is in fact qualified to offer the opinions he or she is proffering; (2) whether each proposed opinion is based upon reliable data and methodology; and (3) whether the proposed testimony would be helpful to the trier of fact or to answer the factual question presented." Deutsche Bank, 2013 WL 5815472, at *13. The party seeking to introduce an expert's opinion bears the burden of establishing, by a preponderance of the evidence, that it is admissible. Daubert, 509 U.S. at 593 n.10; Fed. R. Evid. 702, advisory committee's note (2000). Nevertheless, the Rule 702 inquiry is flexible and courts may exercise significant discretion. Daubert, 509 U.S. at 594; Lees v. Carthage College, 714 F.3d 516, 518 (7th Cir. 2013).
B. Establishing Reliance in an Exchange Act Case
Section 10(b) of the Exchange Act and Rule 10b-5 prohibit making any material misstatement or omission in connection with the purchase or sale of any security. See Halliburton Co. v. Erica P. John Fund, Inc., 134 S. Ct. 2398, 2407 (2014). The elements of an Exchange Act claim are "falsehood in connection with the purchase or sale of securities, scienter, materiality, reliance, causation, and loss." Schleicher v. Wendt, 618 F.3d 679, 682 (7th Cir. 2010). The reliance element is essential to establish a proper connection between a defendant's misrepresentation and a plaintiff's injury. Halliburton, 134 S. Ct. at 2407 (citation omitted); Schleicher, 618 F.3d at 682. "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—e.g., purchasing common stock—based on that specific misrepresentation." Halliburton, 134 S. Ct. at 2407 (citation omitted).
In 1988, the Supreme Court recognized that requiring this direct proof of reliance "would place an unnecessarily unrealistic evidentiary burden on the Rule 10b-5 plaintiff who has traded on an impersonal market." Basic Inc. v. Levinson, 485 U.S. 224, 242 (1988). "That is because, even assuming an investor could prove that he was aware of the misrepresentation, he would still have to show a speculative state of facts, i.e., how he would have acted if the misrepresentation had not been made." Halliburton, 134 S. Ct. at 2407 (citation omitted). The Court also acknowledged that if every plaintiff had to prove direct reliance on the defendant's misrepresentation, individual issues would overwhelm common ones, making class certification under Rule 23(b)(3) inappropriate. Basic, 485 U.S. at 242; see Halliburton, 134 S. Ct. at 2408.
To address these concerns, the Supreme Court "endorsed a fraud-on-the-market' theory, which permits securities-fraud plaintiffs to invoke a rebuttable presumption of reliance on public, material misrepresentations regarding securities traded in an efficient market." Amgen Inc. v. Connecticut Ret. Plans & Trust Funds, 133 S. Ct. 1184, 1188 (2013) (citing Basic, 485 U.S. at 241-49). The fraud-on-the-market theory presumes that "the market price of shares traded on well-developed markets reflects all publicly available information, and, hence, any material misrepresentations." Basic, 485 U.S. at 246; see Schleicher, 618 F.3d at 682 (the fraud-on-the-market theory "supplants 'reliance' as an independent element by establishing a more direct method of causation"). An investor "who buys or sells stock at the price set by [an efficient] market does so in reliance on the integrity of that price—the belief that it reflects all public, material information." Halliburton, 134 S. Ct. at 2408 (citation omitted). Thus, whenever an investor trades in an efficient market, his "reliance on any public material misrepresentations . . . may be presumed for purposes of a Rule 10b-5 action." Basic, 485 U.S. at 247.
Economists have long debated "about the degree to which the market price of a company's stock reflects public information about the company." Halliburton, 134 S. Ct. at 2410. Indeed, the Basic Court acknowledged the debate but declined to weigh in, declaring that it "need not determine by adjudication what economists and social scientists have debated through the use of sophisticated statistical analysis and the application of economic theory." Basic, 485 U.S. at 246 n.24. In recognizing the presumption of reliance, the Supreme Court declined to "adopt any particular theory of how quickly and completely publicly available information is reflected in market price." Id. at 248 n.28. Instead, the presumption of reliance is based on the "modest premise," Halliburton, 134 S. Ct. at 2410, that "market professionals generally consider most publicly announced material statements about companies, thereby affecting stock market prices," Basic, 485 U.S. at 247 n.24. "'That the price of a stock may be inaccurate does not detract from the fact that false statements affect it, and cause loss,' which is 'all that Basic requires.'" Halliburton, 134 S. Ct. at 2410 (favorably quoting Schleicher, 618 F.3d at 685) (alterations omitted).
In 1989, the District of New Jersey became one of the first courts post-Basic to consider what factors should be considered to determine if market efficiency is present. Cammer v. Bloom, 711 F. Supp. 1264 (D.N.J. 1989). In Cammer, the court identified the following factors to evaluate when considering whether a stock trades in an efficient market: (1) average weekly trading volume during the class period; (2) number of security analysts who followed and reported on the stock during the class period; (3) number of market makers; (4) whether the company was entitled to file an S-3 Registration Statement; and (5) whether empirical facts demonstrate a cause-and-effect relationship between unexpected corporate events or financial releases and an immediate response in the stock price. 711 F. Supp. at 1285-87. Since then, a number of other courts, including courts in this district, have adopted the Cammer factors. See, e.g., In re Northfield Labs., Inc. Sec. Litig., 267 F.R.D. 536, 545-46 (N.D. Ill. 2010); Tatz v. Nanophase Technologies Corp., Case No. 01 C 8440, 2003 WL 21372471, at *7 (N.D. Ill. June 13, 2003); Greenberg v. Boettcher Co., 755 F. Supp. 776, 782 (N.D. Ill. 1991). Although the Seventh Circuit has not explicitly adopted the Cammer factors, it agrees that a presumption of reliance under the fraud-on-the-market theory exists for a large, well-followed firm, whose stock trades in a liquid market. Schleicher, 618 F.3d at 682-84; cf. Unger v. Amedisys Inc., 401 F.3d 316, 322 (5th Cir. 2005) ("In many cases, where heavily-traded or well-known stocks are the target of suits, market efficiency will not even be an issue."). Thus, to determine whether a stock trades in an efficient market, experts conduct a Cammer factor analysis. (See Tr. 8-10, 123-25) (Drs. Feinstein and Gompers agreeing that experts employ a Cammer factor analysis to assess market efficiency for purposes of securities litigation).
C. Dr. Feinstein's Opinions
Plaintiff relies on the opinions and testimony of Dr. Feinstein to establish that Groupon traded in an efficient market during the two-month Subclass period. Pursuant to the practice in his field and in order to determine whether Groupon stock traded in an efficient market, Dr. Feinstein testified that he performed a Cammer factor analysis. (Tr. 12). As to the first factor, he observed that on average 2.3 million shares traded daily during the Subclass period with 1.8% of outstanding shares turned over weekly. (Id. 13-14). Dr. Feinstein concluded that these figures demonstrated substantial evidence in favor of finding market efficiency. (Id. 15). As to the second factor, Dr. Feinstein concluded that the combination of 105 buy-side major institutions investing in Groupon stock and 17 sell-side analysts covering the stock constituted "very strong evidence" of market efficiency. (Id. 15-17). In regards to the third factor, because Groupon was listed on the NASDAQ exchange, the market for Groupon stock was "presumptively efficient." (Id. 18); (see Dkt. 254 at 3) (District Judge concluding that because Groupon's stock traded on NASDAQ, "it is undeniably a frequently traded stock in an efficient market"). Moreover, Dr. Feinstein identified approximately 100 market makers willing to buy and sell Groupon stock. (Id.). Thus, he opined that there was "no market-making impediment to the efficiency of Groupon [stock]." (Id. 18-19).
The fourth Cammer factor is whether the company is eligible for S-3 registration. (Tr. 19). Companies permitted by the SEC to file an S-3 Registration Statement are those which meet the $75 million market capitalization requirement and have filed reports with the SEC for twelve consecutive months. 17 C.F.R. § 239.13; see In re Xcelera.com Sec. Litig., 430 F.3d 503, 512 n.9 (1st Cir. 2005). Dr. Feinstein testified that while Groupon had not yet met the twelve-month minimum period of filing SEC reports (the Subclass period is within six months of Groupon filing its IPO), Groupon satisfied the remaining S-3 eligibility requirements. (Id. 19, 21). He calculated Groupon's market capitalization at $800 million, which is over 10 times the $75 million amount needed for S-3 registration. (Id. 19).
Experts use an event study to satisfy the fifth Cammer factor, which seeks a cause-and-effect relationship between the release of new information and movements in the stock price. (Tr. 22); (see id. 125) (Dr. Gompers agreeing that experts use event studies to evaluate the fifth Cammer factor); see also Halliburton, 134 S. Ct. at 2415 (Event studies are "regression analyses that seek to show that the mar- ket price of the defendant's stock tends to respond to pertinent publicly reported events."). Dr. Feinstein explained that to test such a relationship, experts run a regression analysis to determine if stock movements are statistically significant in response to new information. (Id. 24). For purposes of his Groupon analysis, Dr. Feinstein looked for "big news days" during the Subclass period by identifying significant, unexpected earnings-related news. (Id. 26). He further testified that using such dates in event studies is supported by "seminal articles in the literature." (Id.). After reviewing all of the company-specific news in the seven-week Subclass period, only two dates met Dr. Feinstein's objective criteria. (Id. 28). He acknowledged that having only two event dates is "on the low side," but that it was not uncommon for a less-than-two-month Subclass period. (Id.). The first date he picked was February 8, 2012, when, after the close of trading on that day, Groupon reported a loss instead of the profit expected by analysts. (Id. 29, 35). On February 9, 2012—the first trading day after the February 8 earnings miss—Groupon stock fell 15%. (Id. 35). On the second "big news" day, March 30, 2012, Groupon disclosed, after the close of trading, that their previously announced fourth quarter 2011 earnings had been overstated. (Id. 29). On the first trading day following this announcement, April 2, 2012, the stock fell 19%. (Id. 35). Dr. Feinstein testified that there was no other earnings-related news of similar importance during the Subclass period. (Id. 26-30).
A regression analysis is a statistical tool used to understand the relationship between or among two or more variables. Federal Judicial Center, Reference Manual on Scientific Evidence [hereinafter Reference Manual] 305 (3d ed. 2011). In most scientific work, the level to obtain a statistically significant result is set at a 5% level of significance, i.e., that there is no more than a 5% chance that the observed relationship is purely random. Id. 320; see Dean v. China Agritech, No. 11 C 1331, 2012 WL 1835708, at *7 (C.D. Cal. May 3, 2012); United States v. Delaware, No. 01 C 020, 2004 WL 609331, at *10 n.27 (D. Del. March 22, 2004). In other words, a 5% significance level equates to being 95% confident in the observed relationship. See Reference Manual 284-85, 352.
As noted above, the Subclass period runs from February 9 through March 30, 2012.
When he ran a regression analysis on these two dates, the results were highly significant. (Id. 36-37). The relationship between the February 8 earnings announcement and the stock movement on February 9 was statistically significant at a 99.28% confidence level, exceeding the 95% confidence level generally used as a threshold for statistical significance. (Id. 36-37). And the relationship between the March 30 revised earnings announcement and the April 2 stock price decline exceeded a 99.99% confidence level. (Id. 37).
Consistent with literature in his field, Dr. Feinstein used a one-year control period to control for standard volatility. (Tr. 31-32). Nevertheless, varying the control period, according to Dr. Feinstein, made no difference in the results. (Id. 33) ("The event date returns were so big in absolute magnitude, the drops [in stock prices] were so large, that it wouldn't make any difference whatsoever among any reasonable choice of control periods you picked . . . ."). Dr. Feinstein concluded that this empirical information "confirms a cause-and-effect relationship between information and movements in [Groupon] stock." (Id.). In sum, after reviewing all the Cammer factors and conducting an event study, Dr. Feinstein opined that Groupon stock traded in an efficient market during the Subclass period. (Id. 38).
D. Dr. Gompers's Opinions
Dr. Gompers, Defendants' expert, did not dispute Dr. Feinstein's conclusions that (1) the NASDAQ exchange—on which Groupon shares traded—was a well- developed exchange on which most company's stocks traded efficiently most of the time; (2) Groupon shares traded with heavy volume during the Subclass period; (3) weekly turnover in Groupon shares created a "substantial presumption of market efficiency"; (4) equity analysts extensively covered Groupon shares, which is an indicia of market efficiency; and (5) over 100 market makers facilitated trading of Groupon shares. (Tr. 125-30, 169). Nevertheless, Dr. Gompers concluded that Dr. Feinstein did not use an objective scientific approach to determine that Groupon shares traded in an efficient market during the Subclass period. (Id. 80). Dr. Gom-pers opined that Dr. Feinstein used a flawed methodology and implemented it in a flawed manner such that it could not support his market efficiency conclusion. (Id.).
E. Analysis
Defendants do not contest Dr. Feinstein's qualifications. Instead, they contend that Plaintiff has failed to meet his burden of establishing that Dr. Feinstein's "methodology meets the 'exacting standards of reliability' established by the Supreme Court." (Dkt. 273 at 2). Specifically, Defendants argue that Dr. Feinstein (1) employed a flawed event-study methodology, (2) failed to account for short-selling constraints affecting Groupon shares during the Subclass period, and (3) miscalculated Groupon's stock float. (Id.).
1. Event Study Methodology
Defendants contend that Dr. Feinstein used a flawed event study methodology. (Dkt. 273 at 4-10). They assert that Dr. Feinstein failed to (1) adopt an objective and reliable approach to selecting news days, (2) account for confounding infor- mation on news days, (3) account for significant price changes in the absence of news, and (4) select a scientifically sound control period. (Id.).
a. Flawed Selection of News Days
Defendants argue that Dr. Feinstein's selection of only two news days—which bookended the Subclass period—on which to run a regression analysis was subjective and reflected Dr. Feinstein's knowledge of the Subclass period before he started his study. (Dkt. 273 at 4-7). Dr. Feinstein testified that he pre-selected objective criteria—announcements of significant, unexpected earnings-related news—before searching for dates. (Tr. 26). According to Dr. Feinstein, such earnings surprises are objectively determinable by comparing expected versus actual financial results. (Id.) ("an earnings surprise where the market is expecting one number for earnings and a different number comes in, that's always—always big news"). According to Dr. Feinstein, earnings-surprise dates are commonly used by experts conducting event studies. (Id. 26-27). Dr. Feinstein testified that only the two dates studied met this criteria. (Id. 28-30). While having only two dates is not ideal, it was not unusual given the very short Subclass period—less than two months—at issue here. (See id. 28). And the relationship between the earnings surprises and the stock movements on these two dates was statistically significant at a greater than a 99% confidence level. (Id. 37). That the two news days appeared at the beginning and end of the Subclass period does not alter the finding that empirical facts demonstrated a cause-and-effect relationship between unexpected earnings reports and an immediate response in the stock price. See Cammer, 711 F. Supp. at 1285-87.
Dr. Gompers testified that Dr. Feinstein did not consider whether there was an ex-ante earnings expectation prior to the earnings announcement on the two news days. (Tr. 156-57). However, Dr. Feinstein's expert report clearly compared consensus pre-announcement estimates to the reported earnings. (Dkt. 189, Ex. 1 at 19, 23). Dr. Gompers also testified that Dr. Feinstein should have considered two other news days in his analysis. (Tr. 108-09). Dr. Gompers asserted that Dr. Feinstein should have included February 27, 2012, when Groupon announced that it had reached record levels of mobile use, and March 12, 2012, when U.K. regulators announced potential actions against Groupon's advertising claims. (Id. 109-10). But these other two news days did not meet Dr. Feinstein's objective criteria. Neither February 27 nor March 12, 2012, were earnings releases or even mentioned earnings. (Dkt. 275 at 10 & n.5).
Defendants argue that the court's decision in Brown v. China Integrated Energy Inc., No. 11 CV 2559, 2014 U.S. Dist. LEXIS 117764 (C.D. Cal. Aug. 4, 2014), to exclude plaintiff's expert demands the same result here. (Dkt. 273 at 6-7). However, in Brown, the expert "never articulate[d] any objective criteria that he used to identify the events used in his study." 2014 U.S. Dist. LEXIS 117764, at *21. In fact, the expert admitted choosing dates by seeing "how the market reacted to [the infor-mation]—whether the share price jumped or dropped after the information was released." Id. at *24. Here, to the contrary, according to his testimony, Dr. Feinstein first selected the news dates based on objective criteria (unexpected earnings-related reports) before studying any effect on price. (Tr. 25-26).
Likewise, Defendants reliance on Deutsche Bank is inapposite. Defendants contend that like the plaintiff's expert in Deutsche Bank, Dr. Feinstein's selection of event days was "haphazard and inconsistent." (Dkt. 273 at 7). But in Deutsche Bank, the court excluded plaintiff's expert partly because the expert chose all of the defendant's earnings disclosure dates during the class period as event days, regardless of whether the earnings results were unexpected. 2013 WL 5815472, at *2, 11, 16. Here, as discussed above, Dr. Feinstein selected only unexpected, earnings-related news days as event days.
b. Confounding Information
Defendants contend that Dr. Feinstein failed to take into consideration confounding information on February 8, 2012. (Dkt. 273 at 8); see In re Xcelera.com, No. 00 CV 11649, 2008 WL 7084626, at *2 (D. Mass. April 25, 2008) (excluding expert where he failed to consider the risk that stock would be declared a Foreign Personal Holding Company which presented a tax risk to stock holders). Dr. Gompers testified that a few analysts viewed the February 8 earnings announcement in a positive light, thus contradicting Dr. Feinstein's claim that the earnings announcement was unexpected. (Tr. 111). Defendants argue that Dr. Feinstein's failure to take this confounding information into account "certainly draws into question" whether there was efficient price reaction to the information released. (Dkt. 273 at 8; see Tr. 111-12). But Dr. Feinstein credibly testified that by his pre-established objective criteria, Groupon unambiguously failed to meet earnings expectations on February 8. (Tr. 67). As he testified, that a few analysts sought to put a positive spin on the re- sults does not constitute confounding information and does not change the fact that the earnings results were significantly below expectations. (Id.).
c. Absence of News
Defendants argue that Dr. Feinstein failed to determine whether there were statistically significant Groupon stock price changes in the absence of news during the Subclass period. (Dkt. 273 at 8-9). For example, Dr. Gompers asserts that Dr. Feinstein should have considered why Groupon's common stock declined 7.8% on February 13, 2012, in the absence of any news. (Tr. 112-13; see Dkt. 273 at 8). But as Dr. Feinstein testified, to consider this date just because it had a large price drop would be a violation of all the rules of an event study. (Tr. 71). "To first look at the stock price movement and then try to explain it with news is a violation of the event study rules and a violation of scientific principles." (Id.). Moreover, without running a regression analysis, which neither Dr. Feinstein nor Dr. Gompers performed, the stock drop on February 13 is simply not relevant. (Id. 23). The movement on that day could be a "result of the overall stock market," or the sector in which Groupon operates "might be buffeting the stock up and down." (Id. 23-24).
The case cited by Defendants is inapposite. The court in George v. China Auto. Sys., Inc., No. 11 CV 7533, 2013 WL 3357170, at *12-13 (S.D.N.Y. July 3, 2013), excluded the expert because only seven of the sixteen news days identified by plaintiff's expert resulted in a statistically significant market reaction. Moreover, six of the ten largest stock movements occurred on non-news days. These results led the George court to conclude that the plaintiff failed to satisfy his burden of establishing market efficiency. Id. Here, all of the news days identified by Dr. Feinstein resulted in statistically significant price reactions and, arguably, only one non-news date included a sizeable price change.
The other cases cited by Defendants do not involve market efficiency or fraud-on-the-market presumptions. See Remien v. EMC Corp., No. 04 C 3727, 2008 WL 597439 (N.D. Ill. March 3, 2008) (employment discrimination class action); Bolden v. Walsh Group, No. 06 C 4104, 2012 WL 1079898 (N.D. Ill. March 30, 2012) (same).
d. Control Period
Defendants contend that Dr. Feinstein's event study failed to select a control period that represented normal Groupon stock price movement. (Dkt. 273 at 9-10). Dr. Gompers testified that a normal trading period is needed "to get an estimate of how much of Groupon stock price movement is explained by the market and how much is explained by the industry." (Tr. 113). Defendants argue that Dr. Feinstein's selection of the control period both before and after the Subclass period was faulty. (Dkt. 273 at 9).
Dr. Feinstein testified that consistent with the literature of his field, he used a one-year control period—from the time of the IPO in November 2011 through November 2012—to control for volatility. (Tr. 31-32, 113). If a longer period is used, he testified, staleness may result. (Id. 31-32). Dr. Gompers testified that because the period of time after the IPO is a "quiet period when the underwriting analyst can't issue reports," it is not considered a normal trading period. (Id. 114). Dr. Feinstein acknowledged that the literature in his field suggests that the day after the IPO should be excluded, which he did. (Id. 32). Moreover, Dr. Feinstein concluded that if he were to exclude a longer quiet period, it would have no impact on the result. (Id. 33). "The event dates were so big in absolute magnitude, the drops were so large, that it wouldn't make any difference whatsoever among any reasonable choice of control periods you picked." (Id.).
Defendants also criticize Dr. Feinstein's use of dummy variables in the control period after March 30, 2012. (Dkt. 273 at 10). Dr. Feinstein explained that dummy variables are widely used during a control period to extract atypical days. (Tr. 33). Consistent with other experts, he used dummy variables to control for potentially abnormal returns related to earnings announcements. (Id. 33-34). Accordingly, he used dummy variables during the control period for all earnings-related trading dates, including the event days. (Id.). He also used a dummy variable for May 14, 2012, because it was not only an unusual trading day, but also because it was contemporaneously described as atypical by the news media. (Id. 34). Dr. Gompers testified that while the use of dummy variables is appropriate, it must be done with an objective approach. (Id. 115). Dr. Gompers specifically criticized the use of a May 14, 2012 dummy variable, contending that its use on that day was "backward looking." (Id. 116).
The Court is not persuaded. Dr. Feinstein used only five dummy variables during the control period, which is on the low side. (Tr. 34). Dr. Feinstein explained that he used a May 14 dummy variable because the news media described the day as atypical and it appeared there was news leakage on the day before the May 15, 2012 earnings announcement. (Tr. 34). In any event, Dr. Feinstein testified that including the May 14 dummy variable "made no difference at all" to his results. (Id.); (see also id. at 161) (Dr. Gompers did not test whether excluding the May 14 dummy variable would have changed Dr. Feinstein's results).
Dummy variables were used on February 9, April 2, May 14, May 15, and August 14, 2012. (Dkt. 189, Ex. 1 at 59).
Defendants, relying on In re Xcelera.com and In re Northfield Labs, argue that Dr. Feinstein's opinion is unreliable because of his use of dummy variables for five days during the year-long control period. In Northfield, Judge Marovich excluded an expert who used 117 dummy variables over a five-year class period:
in an event study, the author of the study attempts to determine the usual volatility of the stock by excluding a few key dates on which the stock price is expected to react to news. For example, the author would exclude (from the calculation of the stock's usual volatility) the four dates of the year when the company issued its quarterly earnings, because on those dates, the share price is expected to change based on new information. In Dr. Hakala's study, instead of excluding a few dates, he excludes 117 event dates.267 F.R.D. at 548. Similarly, in Xcelera.com, the court faulted plaintiff's expert for using dummy variables in over one third of the trading days in the class period and for failing to include a day where the company made a corrective earnings disclosure. 2008 WL 7084626, at *1. The Court is not persuaded by these cases since Dr. Feinstein used only five dummy variables during a year-long control period and used dummy variables for all earnings-related trading days during the control period. (Tr. 28-30, 34).
f. Summary
The Court finds that Plaintiff has established by a preponderance of evidence that Dr. Feinstein's event study methodology met Daubert's exacting standards. He relied on principles and methods generally accepted by market efficiency experts. He correctly analyzed "market efficiency" from the perspective of whether unexpected information quickly affected Groupon stock prices—not whether the price of Groupon stock accurately reflected all information. See Halliburton, 134 S. Ct. at 2410. Dr. Feinstein determined with a very high degree of statistical confidence that a cause-and-effect relationship existed between surprise corporate announcements and an immediate response in Groupon's stock price. Basic, 485 U.S. at 246; Cammer, 711 F. Supp. 2d at 1285-87.
2. Short-Selling Constraints
Defendants argue that Dr. Feinstein failed to consider short-selling constraints that affected Groupon shares during the Subclass period. (Dkt. 273 at 10-13). Dr. Gompers noted that "short selling is one of the fundamental cornerstones of helping to facilitate an efficient market through . . . arbitrage, through the ability of those with negative information or negative views to trade on that information." (Tr. 88). He explained that "if short selling is constrained, then people with negative views of a stock can't trade, so a stock price might react too much to positive information . . . [or] it may under react to negative news." (Id. 91). Dr. Gompers opined that a high lending fee on shorting Groupon stocks during the Subclass period rendered the market constrained. (Id. 89). He concluded that by failing to evaluate short selling constraints, Dr. Feinstein's event study was flawed. (Id. 90-91).
"A short sale takes place when a seller, believing the price of a stock will fall, borrows stock from a lender and sells it to a buyer. Later, the seller buys similar stock to pay back the lender, ideally at a lower price than he received on the sale to the buyer." Federal Judicial Center, Federal Securities Law [hereinafter Federal Securities] 171 (3d ed. 2011). As Dr. Gompers explained, if an investor identifies negative information but doesn't own the stock, the only way to get that information into the market is for the investor to short sell the stock. (Tr. 88).
Dr. Feinstein agreed that the ability to short shares enhances market efficiency. (Tr. 52). Moreover, he acknowledged that if short selling is expensive, it could be a factor weighing against market efficiency. (Id. 53). In this case, Dr. Feinstein determined that short selling Groupon stock was possible, that there was an active market for short selling Groupon stock, and that there was no regulatory constraint or ban on short selling Groupon stock. (Id. 53-54); cf. Deutsche Bank, 2013 WL 5815472, at *12 (complete ban on short selling is a factor weighing against finding market efficiency).
The Court is satisfied that Dr. Feinstein's methodology satisfies Daubert. While Dr. Gompers concern about the high cost of short selling Groupon shares during the Subclass period may suggest that the market for Groupon shares was not perfectly efficient, Dr. Feinstein's methodology established that the price of Groupon stock was affected by false statements, which is all that Basic requires. See Halliburton, 134 S. Ct. at 2410. In any event, there is no evidence that the high cost of short selling Groupon shares affected the market. Dr. Gompers acknowledged on cross-examination that during the Subclass period, (1) millions of shares shorted every day, (2) Groupon shares were shorted both before and during the Subclass period, (3) short interest grew during the Subclass period, (4) traders were able to bet against Groupon in the options market if they desired, and (5) no actual traders were identified who were unable to short Groupon stock. (Tr. 123, 132, 134, 139, 142).
Defendants also criticize Dr. Feinstein's methodology for failing to account for constraints in the put-call market. (Dkt. 273 at 12; Dkt. 279 at 10). "A call option is a contract between a seller (the option writer) and a buyer under which the option buyer has the right to exercise the option and thereby purchase the underlying security at an agreed-on price." Federal Securities 167. "A put option gives the option's buyer the right to exercise the option by selling the underlying security. The put-option seller must purchase the underlying security at the agreed-on price (the strike price) if the option is exercised on or before the expiration date." Id. 169. Dr. Gompers testified that put-call parity is an indication that the short sale market is not working efficiently. (Tr. 96, 104). Again, while these theoretical constraints may implicate the degree to which Groupon stock prices accurately reflect public information, a put-call analysis is simply not required to establish market efficiency under Basic and Halliburton. Dr. Feinstein's methodology is not undermined by his failure to perform unnecessary analyses. See Daubert, 509 U.S. at 597 (district court has the responsibility to ensure "that an expert's testimony both rests on a reliable foundation and is relevant to the task at hand"). Moreover, Dr. Gompers conceded that his analysis used artificial prices in a manner that could not actually be performed in the market. (Tr. 146-48).
3. Stock Float/Market Capitalization
Defendants contend that the reliability of Dr. Feinstein's analysis is undermined by his error in calculating Groupon's stock float. (Dkt. 273 at 13-15). The fourth Cammer factor analyzes whether a company is eligible for S-3 registration under SEC rules, which require 12 months of periodic reporting and a market capitalization exceeding $75 million in value. (Tr. 19). Dr. Feinstein testified that Groupon could not satisfy S-3 eligibility because it was a new public company with less than 12 months of periodic filings. (Id. 21). He concluded that while the fourth Cammer factor "wasn't evidence in favor of market efficiency, . . . it didn't seem to be evidence against market efficiency either." (Id.).
Stock float is the total number of shares held by the public, rather than insiders. Krogman v. Sterritt, 202 F.R.D. 467, 478 (N.D. Tex. 2001).
Market capitalization is equal to the stock float times the share price. Krogman, 202 F.R.D. at 478. During the relevant time period, Groupon's stock price was about $20 per share. (Tr. 19, 47-48).
Both Dr. Feinstein and Dr. Gompers agree that Groupon's market capitalization during the Subclass period substantially exceeded the $75 million needed for S-3 registration. (Tr. 20, 46-48, 85). They differed on whether to include shares in the float that were subject to an underwriter lockup provision. (Id. 20, 85). In his expert report, Dr. Feinstein opined that Groupon's stock float was 302 million shares, resulting in a market capitalization of between $5.05 and $6.39 billion. (Dkt. 189, Ex. 1 at ¶ 61). Dr. Gompers disputed Dr. Feinstein's opinion, asserting that Groupon's stock float during the Subclass period was at most 40.25 million shares because the remaining shares were subject to lockup provisions and were not available for trade. (Dkt. 212, Ex. A at ¶¶ 38-39).
On cross-examination, Dr. Feinstein conceded that the number of shares actually available for trading during the Subclass period was only 40.25 million. (Tr. 43). Defendants argue that Dr. Feinstein's "error critically undermines the reliability of his entire market efficiency analysis." (Dkt. 273 at 13). The Court is not persuaded. Even assuming that the lockup shares should have been excluded, Groupon's market capitalization was still $800 million, far exceeding the $75 million threshold. (See Dkt. 275 at 7; Tr. 46-48). And given that the fourth Cammer factor was otherwise not met because Groupon was a new public company, any error by Dr. Fein- stein is not significant enough to undermine the overall methodology of his market efficiency analysis.
Defendants argue that Dr. Feinstein's float calculation error exacerbated his failure to account for short selling constraints in Groupon stock during the Subclass period. (Dkt. 273 at 14-15; Dkt. 279 at 4). Dr. Gompers testified that "the shares which are in the float are what determines how the arbitrage market works, how the short sale market works." (Tr. 86). But as discussed above, there is no evidence of any actual short sale constraints during the Subclass period. (Id. 123, 132, 134, 139, 142). And, more importantly, these theoretical constraints do not undermine Dr. Feinstein's methodology for determining whether there was an efficient market for Groupon stock during the Subclass period.
--------
4. Other Reasons
Defendants also suggest that the Court should reject Dr. Feinstein's analysis here because other courts have found his opinions "unreliable." (Dkt. 273 at 3-4). But the Court's gatekeeping role under Daubert mandates that it take an independent inquiry into the proffered expert's qualifications, data, and methodology. Cf. Deutsche Bank, 2013 WL 5815472, at *12 (A "court should not simply accept a proffered expert—even if he has been accepted as such before; the law requires an inquiry into whether an expert is an expert, and some degree of 'regulation of the subjects and theories about which an expert may testify.'") (quoting Daubert, 509 U.S. at 589). In any event, the cases cited by Defendants are distinguishable. In Finkelstein v. Liberty Digital, No. 19598, 2005 WL 1074364, at *1, 12 (Del. Ch. April 25, 2005), Dr. Feinstein was asked to opine about the fair value of the plaintiffs' shares on the date of a merger involving defendant and another entity. The court rejected Dr. Feinstein's opinions, finding them "unreliable" and described his "assumptions [as] bear[ing] no relationship to reality." Id. at *13-17. But Finkelstein involved a valuation opinion, which has no relevance to the market efficiencies analysis and methodologies used by Dr. Feinstein here. (See Tr. 11-12).
In Dean v. China Agritech, No. 11 CV 1331, 2012 WL 1835708, at *3, 7 (C.D. Cal. May 3, 2012), Dr. Feinstein was asked to opine about the methodology used by another of plaintiff's experts (Dr. Werner) to test market efficiency. Dr. Feinstein also conducted two event studies, finding that "in the aggregate, there was a statistically significant level of correlation between Agritech's disclosures and movement in the price of its stock." Id. at *7. Dr. Feinstein's methodologies were not challenged and his opinions were admitted. Id. at *1, 4; (see Tr. 12, 51-52). Indeed, the court denied defendant's motion to exclude Dr. Werner's opinion, partly based on Dr. Feinstein's opinion supporting Dr. Werner's methodologies. Dean, 2012 WL 1835708, at *3-4. Nevertheless, the court discounted Dr. Finkelstein's event study because it was measured in the aggregate and differed from Dr. Werner's results. Id. at *7. Here, however, Dr. Feinstein's event studies were not measured in the aggregate and no other studies contradicted his results.
5. Summary
At the evidentiary hearing, Dr. Feinstein testified in a clear and credible manner. The Court found him to be reliable and clearly an expert in conducting event studies. The Court finds that Plaintiff has established by a preponderance of evidence that Dr. Feinstein's event study methodology met Daubert's exacting standards. He relied on principles and methods generally accepted by market efficiency experts. His data and methodology targeted the ambit of market efficiency required by Cammer, Basic, and Halliburton. In this regard, Dr. Feinstein remained focused on the correct legal standard for market efficiency, as laid out in Supreme Court rulings. He correctly analyzed "market efficiency" from the perspective of whether unexpected information quickly affected Groupon stock prices—not whether the price of Groupon stock accurately reflected all information. See Halliburton, 134 S. Ct. at 2410. Dr. Feinstein determined with a very high degree of statistical confidence that a cause-and-effect relationship existed between surprise corporate announcements and an immediate response in Groupon's stock price. Basic, 485 U.S. at 246; Cammer, 711 F. Supp. 2d at 1285-87. Thus, Dr. Feinstein's event study methodology not only "rests on a reliable foundation" but is also directly "relevant to the task at hand." Daubert, 509 U.S. at 597.
While the Court found Dr. Gompers credible and knowledgeable, and clearly an expert on efficient capital markets, his attempts to undermine Dr. Feinstein's opinions were not persuasive. Defendants' arguments suggest that the fraud-on-the-market presumption requires a perfectly efficient market. But this argument was squarely rejected by the Halliburton Court. 134 S. Ct. at 2410 ("Debates about the precise degree to which stock prices accurately reflect public information are thus largely beside the point.") (emphasis in original); see In re Xcelera.com, 430 F.3d at 510 ("the fraud-on-the-market presumption of reliance does not depend on the accuracy of the market price, and whether it mirrors the best possible estimates, in light of all available information, of the actual economic values of securities in terms of their expected risks and returns") (citation and alterations omitted).
The Court need not adjudicate the debate between Dr. Feinstein and Dr. Gom-pers "about the degree to which the market price of [Groupon's] stock reflects public information about the company—and thus the degree to which an investor can earn an abnormal, above-market return by trading on such information," in order to determine market efficiency under Basic. Halliburton, 134 S. Ct. at 2410. Indeed, that the price of Groupon stock during the Subclass period may not have perfectly reflected all public information "is largely beside the point." Id. Dr. Gompers was unable to contradict Dr. Feinstein's opinions that alleged false statements affected the price of Groupon stock and caused losses. As another district court observed in rejecting a similar argument made by Dr. Gompers, Defendants "rely on factors that are not legally relevant." Lumen v. Anderson, 280 F.R.D. 451, 460 (WD. Mo. 2012). "This does not mean that Defendants (or their expert, Paul Gompers) are incorrect in what they say—it means that Defendants (and their expert) often describe a different conception of an efficient market than is used by the law." Id.
The Court concludes that Dr. Feinstein's opinions and testimony are based upon reliable data and methodology. See Deutsche Bank, 2013 WL 5815472, at *13. Plaintiff has met his burden of establishing, by a preponderance of the evidence, that Dr. Feinstein's opinions and testimony are admissible. Daubert, 509 U.S. at 593 n.10; Fed. R. Evid. 702; see Lees, 714 F.3d at 518.
III. CONCLUSION
For the reasons stated above, Defendants' Motion to Exclude the Opinions and Testimony of Steven P. Feinstein [234] is DENIED. Dated: March 5, 2015
ENTER:
/s/_________
MARY M. ROWLAND
United States Magistrate Judge