Opinion
Civil No. 08cv1689 AJB (RBB).
2012-12-14
In re NOVATEL WIRELESS SECURITIES LITIGATION. This Document Relates to All Actions.
Douglas R. Britton, Matthew P. Montgomery, Robert R. Henssler, Jr., Robbins Geller Rudman & Dowd LLP, San Diego, CA, for Plaintiffs. Cary D. Sullivan, Eric Landau, Richard J. Grabowski, Travis Biffar, Jones Day, Irvine, CA, Meir Feder, Jones Day, New York, NY, Patricia J. Villareal, Thomas Ray Jackson, Jones Day, Dallas, TX, Robert S. Brewer, Jr., Jones Day, San Diego, CA, Kevin Hugh Logan, Jones Day, Washington, DC, for Defendants.
Douglas R. Britton, Matthew P. Montgomery, Robert R. Henssler, Jr., Robbins Geller Rudman & Dowd LLP, San Diego, CA, for Plaintiffs. Cary D. Sullivan, Eric Landau, Richard J. Grabowski, Travis Biffar, Jones Day, Irvine, CA, Meir Feder, Jones Day, New York, NY, Patricia J. Villareal, Thomas Ray Jackson, Jones Day, Dallas, TX, Robert S. Brewer, Jr., Jones Day, San Diego, CA, Kevin Hugh Logan, Jones Day, Washington, DC, for Defendants.
ORDER (1) DENYING DEFENDANTS' MOTION TO EXCLUDE EXPERT TESTIMONY OF BJORN STEINHOLT AND (2) AMENDING PLAINTIFFS' COMPLAINT
[Doc. No. 299]
ANTHONY J. BATTAGLIA, District Judge.
Defendants filed a motion to exclude testimony of Plaintiffs' expert Bjorn Steinholt on February 15, 2011. (Doc. No. 299.) Based upon the moving papers and arguments and for the reasons set forth herein, Defendants' motion is hereby DENIED for the reasons set forth below.
Background
Plaintiffs allege that between February 27, 2007, and November 10, 2008 (the “Class Period”), Defendants engaged in a fraudulent scheme to inflate Novatel's stock value so that Defendants could sell their stock in the company for a profit. (FAC, Doc. No. 23 at ¶¶ 1, 12). Plaintiffs contend that Novatel's success was largely dependent on its ability to supply wireless modems to its two largest customers, Sprint and Verizon, which in 2006 accounted for 38.2% and 19.7% of Novatel's revenue respectively. ( Id. at ¶ 14). According to Plaintiffs, Defendants “knew that the market was particularly sensitive to information about these customers” and “[s]trong financial results would surely spur an increase in Novatel's stock price whereas any negative information regarding these customers would reduce it.” ( Id. at ¶ 14.)
Plaintiffs allege that throughout the Class Period, Defendants Weinert and Leparulo misrepresented the financial condition of the Company because they told investors that the Company was seeing strong demand for its products, and did not disclose to investors that Novatel did not have an adequate “product mix” to meet the needs of its customers. ( Id. at ¶¶ 57(a)(iii), 62(a)(ii), 66(a)(ii), 73(a)(ii).) Plaintiffs also allege that Novatel covered up the slowdown in its business by shipping product “early,” which purportedly violated accounting rules governing revenue recognition. ( Id. at ¶ 6.) Plaintiffs allege that during this time period, Defendants were selling significant amounts of their Novatel holdings. ( Id.) Plaintiffs allege that during the Class Period, Defendants sold 1,258,466 shares of Novatel stock for almost $29 million in proceeds. ( Id. at ¶ 15.) Plaintiffs allege that 62% of the Defendants' Class Period sales occurred in June and July 2007, just before the market learned about these concealed facts. ( Id. at ¶ 16.) Plaintiffs claim that four specific stock price declines on July 20, 2007, February 21, 2008, April 15, 2008, and August 20, 2008, resulted from the market learning of these allegedly concealed facts. ( Id. at ¶¶ 125–28.)
Legal Standard
Federal Rule of Evidence 702 provides that expert testimony is admissible if “scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue.” Fed.R.Evid. 702. Expert testimony under Rule 702 must be both relevant and reliable. Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 589, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993). When considering evidence proffered under Rule 702, the trial court must act as a “gatekeeper” by making a preliminary determination that the expert's proposed testimony is reliable. Elsayed Mukhtar v. Cal. State Univ., Hayward, 299 F.3d 1053, 1063 (9th Cir.2002), amended by319 F.3d 1073 (9th Cir.2003).
As a guide for assessing the scientific validity of expert testimony, the Supreme Court provided a non-exhaustive list of factors that courts may consider: (1) whether the theory or technique is generallyaccepted within a relevant scientific community, (2) whether the theory or technique has been subjected to peer review and publication, (3) the known or potential rate of error, and (4) whether the theory or technique can be tested. Daubert, 509 U.S. at 593–94, 113 S.Ct. 2786;see also Kumho Tire Co., Ltd. v. Carmichael, 526 U.S. 137, 119 S.Ct. 1167, 143 L.Ed.2d 238 (1999). The Ninth Circuit also has indicated that independent research, rather than research conducted for the purposes of litigation, carries with it the indicia of reliability. See Daubert v. Merrell Dow Pharmaceuticals, Inc., 43 F.3d 1311, 1317 (9th Cir.1995) (“Daubert II ”). In particular, using independent, pre-existing research “provides objective proof that the research comports with the dictates of good science” and is less likely “to have been biased by the promise of remuneration.” Id. If the testimony is not based on “pre-litigation” research or if the expert's research has not been subjected to peer review, then the expert must explain precisely how he went about reaching his conclusions and point to some objective source—a learned treatise, the policy statement of a professional association, a published article in a reputable scientific journal or the like, to show that he has followed the scientific method, as it is practiced by (at least) a recognized minority of scientists in his field. Id. at 1318–19 (citing United States v. Rincon, 28 F.3d 921, 924 (9th Cir.1994)); see also Lust v. Merrell Dow Pharmaceuticals, Inc., 89 F.3d 594, 597 (9th Cir.1996). The proponent of the evidence must prove its admissibility by a preponderance of proof. See Daubert, 509 U.S. at 593 n. 10, 113 S.Ct. 2786.
Discussion
Defendants have filed a motion seeking to exclude the testimony of Plaintiffs' loss causation expert, Bjorn Steinholt, as inadmissible under Rule 702 of the Federal Rules of Evidence as well as Daubert and its progeny. (Defs.' Mot., Doc. No. 299–1.) Among other things, Defendants contend primarily that: (1) Steinholt's opinions on loss causation are based on a defective event study, and (2) Steinholt's damages calculations cannot be reconciled with Dura Pharms., Inc. v. Broudo, 544 U.S. 336, 125 S.Ct. 1627, 161 L.Ed.2d 577 (2005), and subsequent Ninth Circuit case law. ( Id.)
A. Steinholt's Opinion Regarding Loss Causation
1. Loss Causation Generally
In order to succeed on a private right of action brought pursuant to Section 10(b) of the Securities and Exchange Act, a plaintiff must prove, inter alia, loss causation. See Stoneridge Inv. Partners, LLC v. Scientific–Atlanta, Inc., 552 U.S. 148, 156–57, 128 S.Ct. 761, 169 L.Ed.2d 627 (2008). Loss causation is the causal connection between a defendant's material misrepresentation and a plaintiff's loss. Dura Pharms., Inc. v. Broudo, 544 U.S. 336, 344–45, 125 S.Ct. 1627, 161 L.Ed.2d 577 (2005); see also In re Oracle Corp. Sec. Litig., 627 F.3d 376, 387 (9th Cir.2010). “A plaintiff bears the burden of proving that a defendant's alleged unlawful act ‘caused the loss for which the plaintiff seeks to recover damages.’ ” In re Gilead Scis. Sec. Litig., 536 F.3d 1049, 1055 (9th Cir.2008) (quoting 15 U.S.C. § 78u–4(b)(4)). In Dura, the Supreme Court held that a person who misrepresents the financial condition of a corporation in order to sell stock is only liable to a relying purchaser for the loss the purchaser sustains when the facts “become generally known” and “as a result” share value depreciates. 544 U.S. at 344–45, 125 S.Ct. 1627. To adequately plead loss causation, the Supreme Court held, a plaintiff must allege that the “share price fell significantly after the truth became known.” Id. at 347, 125 S.Ct. 1627.
To meet this burden, a plaintiff must demonstrate a causal connection between the material misrepresentation and the injury suffered by plaintiff. Dura, 544 U.S. at 342, 125 S.Ct. 1627;see also Metzler Inv. GMBH v. Corinthian Coll., Inc., 540 F.3d 1049, 1063 (9th Cir.2008) (citing In re Daou Systems, Inc., 411 F.3d 1006, 1025 (9th Cir.2005)). A plaintiff is not required to show “that a misrepresentation was the sole reason for the investment's decline in value” in order to establish loss causation. Daou, 411 F.3d at 1025 (citing Robbins v. Koger Props., Inc., 116 F.3d 1441, 1447 n. 5 (11th Cir.1997)). “ “[A]s long as the misrepresentation is one substantial cause of the investment's decline in value, other contributing forces will not bar recovery under the loss causation requirement” but will play a role “in determining recoverable damages.” ” Id. (quoting Robbins, 116 F.3d at 1447 n. 5). However, establishing price inflation alone is insufficient. Dura, 544 U.S. at 342, 125 S.Ct. 1627. A plaintiff must show that an economic loss occurred after the truth behind the misrepresentation or omission became known to the market. Id. Furthermore, the decline in stock price caused by the revelation of that truth must be statistically significant. Id. at 342–47, 125 S.Ct. 1627.
2. Steinholt's Loss Causation Methodology
In order to assess the loss causation aspect of Plaintiffs' case, Steinholt created and relied upon an event study assessing the impact of the fraudulent conduct alleged by Plaintiffs. (Steinholt Rep., Doc. 299–4, ¶ 34.) Event studies analyze the price movement following a disclosure of new information to: “(a) assess statistical significance; and (b) quantify the portion of the price movement not explained by market and/or industry factors, i.e., the company specific portion of the price movement.” ( Id.) When conducting an event study, Steinholt follows this general process: define the event(s); adjust for market and industry factors; select control period; calculate predicted return, abnormal returns and t-statistics; and interpret results. ( Id.) Steinholt analyzes both specific event days containing alleged misrepresentations that materially changed the public mix of information regarding Novatel stock, and specific event days that partially revealed the alleged truth through corrective disclosures. ( Id. at ¶ 35.)
Defendants do not object to Steinholt's event study methodology generally, but contend that the event study is flawed for several reasons. (Defs.' Mot., Doc. No. 299–1 at 11.) First, Defendants contend that Steinholt failed to isolate those portions of stock price decline that are not attributable to the alleged corrective disclosures as opposed to other negative factors affecting Novatel's stock price. ( Id.) Second, Defendants contend that Steinholt should have utilized a two-tailed test rather than a one-tailed test when considering the statistical significance of the stock price decline on July 20, 2007. ( Id. at 17.) Third, Defendants claim that Steinholt improperly manipulated the control period that he used for his event study in order to find statistical significant for the stock price drop on July 20, 2007. ( Id. at 15.) Fourth, Defendants argue that the intraday analysis regarding the stock price decline on July 20, 200 cannot establish loss causation. ( Id. at 20.) Insomuch as Defendants' objections are limited to these aspects of Steinholt's event study, the Court tailors its analysis accordingly.
a. Steinholt's Consideration of Non–Fraud Related Influences on Stock Prices
Here, Plaintiffs allege that Novatel's stock price fell after the truth regarding Defendants' false and misleading statements became known through corrective disclosures on each of the following dates: July 20, 2007; February 20, 2008; April 15, 2008; August 19, 2008; and November 10, 2008. Steinholt's event study analyzes the impact of the alleged corrective disclosures while also adjusting for market and industry factors affecting stock prices. (Steinholt Rep., Doc. No. 299–4, ¶¶ 34, 36.) Nevertheless, Defendants contend that Steinholt's event study is insufficient because it does not distinguish between the alleged corrective disclosures and other company-specific news that may have caused Novatel's stock price to decline. (Defs.' Mot., Doc. 299–1 at 6.) Specifically, Defendants argue that the allegedly corrective disclosures “contained multiple pieces of information that are entirely separate and distinct from Plaintiffs' allegations in this case” that should have been addressed by Steinholt. ( Id. at 12.)
In this situation, Defendants are correct that Steinholt's event study could have included an analysis of non-fraud-related company-specific news. However, the failure to consider one of the relevant variables does not dictate exclusion of the report as unreliable. Generally, the failure to include a variable in a regression analysis affects the probative value of the analysis and not necessarily its admissibility. Bazemore v. Friday, 478 U.S. 385, 400, 106 S.Ct. 3000, 92 L.Ed.2d 315 (1986); see also In re REMEC Inc. Sec. Litig., 702 F.Supp.2d 1202, 1273 (S.D.Cal.2010). “Where a study accounts for the ‘major factors' but not ‘all measurable variables,’ ” it is nevertheless admissible. See REMEC, 702 F.Supp.2d at 1273 (quoting Bazemore, 478 U.S. at 400, 106 S.Ct. 3000 (citations omitted)). Here, Steinholt weighed both market and industry factors in relation to the allegedly fraudulent statement and corrective disclosures. (Steinholt Rep., Doc. No. 299–4, ¶¶ 34, 36.) He also reviewed additional non-fraud related information referenced by Defendants' expert, and opined in his rebuttal that the impact of this information on the loss causation analysis would either be immaterial or inconsistent with the facts Plaintiffs intend to prove at trial. (Pls.' Resp., Doc. 320 at 11 (citing Steinholt's Rebuttal, Doc. No. 299–5, ¶¶ 13, 15, 19, 30).) Specifically, Steinholt noted that he “would have considered any additional information unrelated to the alleged fraud presented by Dr. Tabak if it could be shown, based on sound economic analysis, to have had a material negative impact on Novatel's stock price.” (Rebuttal Doc. No. 299–5, ¶ 25.) For these reasons, while Steinholt's failure to include one non-fraud-related variable may affect the probative value of the event study, it does not require exclusion of the report as unreliable under Daubert.
In REMEC, the court considered a Daubert motion objecting to an event study performed regarding REMEC stock. 702 F.Supp.2d at 1273–75. The court found that a key flaw in the expert's analysis was his failure to separate the loss caused by the disclosure of corrective information from loss caused by the disclosure of other company-specific information. Id. at 1273–1274. The expert also failed to account for industry-specific news, market-specific news, and other measurable macroeconomic variables in his analysis. Id. at 1273. Finding that the regression model suffered from several deficiencies in addition to the omission of key independent variables, the court determined that the expert's study was so incomplete that it became inadmissible as irrelevant and unreliable. Id. Unlike the expert's report in REMEC, Steinholt's event study considers both market and industry factors. Accordingly, Steinholt's alleged failure to consider other company-specific information within Novatel's alleged corrective disclosures is more an issue of probative value than admissibility.
b. Steinholt's Use of a One–Tailed Test to Assess Statistical Significance on July 20, 2007
Within their broader objection to Steinholt's analysis of July 20, 2007, Defendants include a brief argument contending that Steinholt's analysis should be found unreliable because he utilized a onetailed test rather than a two-tailed test to determine statistical significance for July 20, 2007. (Defs.' Mot., Doc. No. 299–1 at 17.) As explained by Defendants' expert, Tabak, “a two-sided test allows for the possibility of finding that a result, in this case a stock price movement, is abnormally positive or negative,” whereas a “one-sided test assumes that the result is meaningful only if it goes in one direction.” (Tabak Decl., Doc. No. 299–12, ¶ 24.) Tabak contends that a one-sided test effectively doubles the error rate of finding a price movement to be statistically significant, and he further notes that financial economists nearly exclusively rely on two-sided tests as opposed to one-tailed tests. ( Id. at ¶¶ 25, 26.) In contrast, Steinholt opines in his report that one-tailed tests are “useful when an event can be independently determined to have a negative future cash flow impact, or positive future cash flow impact, but not both.” (Steinholt Rep., Doc. No. 299–4, ¶ 40 n. 20.) For this reason, Plaintiffs contend that the onetailed test was particularly appropriate because the information disclosed on July 20, 2007 “was information that from an economic point of view could only result in a negative impact on Novatel's stock price.” (Pls.' Resp., Doc. No. 320 at 16 n. 5.)
Neither party provides Ninth Circuit case law regarding the use of a one-tailed test as opposed to a two-tailed test in a securities case. The only legation citation provided is a Fourth Circuit case that found it improper to use a one-tail test of statistical significance because it skews in favor of plaintiffs in employment discrimination cases. (Defs.' Mot., Doc. No. 299–1 at 17 (citing E.E.O.C. v. Fed. Reserve Bank of Richmond, 698 F.2d 633, 655–56, 661 (4th Cir.1983)rev'd on other grounds sub nom. Cooper v. Fed. Reserve Bank of Richmond, 467 U.S. 867, 104 S.Ct. 2794, 81 L.Ed.2d 718 (1984)).) Standing alone the Fourth Circuit's decision considering the use of one-tailed tests specifically with regard to discrimination is not sufficient to find the use of a one-tailed test in a securities case unreliable.
Having considered both experts' views and having found no standard governing the use of one-tailed tests to establish statistical significance of a stock price decline, the Court is faced with a difference of opinion between experts. Simply because one type of test is used more often than another does not necessarily establish that the other test is unreliable. Whereas Dr. Tabak finds the use of the one-tailed test to be unreliable and Steinholt has provided a reasonable explanation for using the onetailed test, the Court finds that it is not an issue of admissibility, but rather of probative value to be addressed at trial. In this instance, Defendants' objection relates more to the appropriate weight to be given to Steinholt's analysis, if any, which is for the trier of fact to consider.
c. Steinholt's Adjustment of the Control Period from One Year to 120 Days
When creating an event study, Steinholt selects “a control period with normal price returns unaffected by the events examined, or estimation window, to determine the normal historical statistical relationship between Novatel's price returns and that of the market and/or industry indices.” (Steinholt Rep., Doc. No. 299–4, ¶ 37.) In this instance, Steinholt utilized both a one-year control period and, subsequently, a 120–day control period when analyzing the impact of the alleged fraud. ( Id.) Defendants contend that Steinholt utilized the 120–day control period in his event study rather than the initial one-year control period in order to achieve a more favorable result for the price decline on July 20, 2007. (Defs.' Mot., Doc. No. 299–1 at 15.) Defendants suggest that Steinholt was dissatisfied with the results of his initial one-year control period which found that the price decline on July 20th was statistically significant at the 90% level of confidence. ( Id. (quoting Steinholt Decl., Doc. No. 299–15, ¶ 40).) Defendants contend that Steinholt recognized that 90% statistical significance was insufficient so he tried a different control period in order to yield higher statistically significant results. ( Id.) The 120–day control period found that the July 20, 2007 stock price decline was statistically significant at the 95% level. ( Id. at 17 (quoting Steinholt Rep., Doc. No. 299–4, at ¶ 43 n. 1).)
In his report, Steinholt provides the following explanation for his decision to calculate loss causation using two different control periods when analyzing the relevant dates:
For my initial analysis of each day during the Class Period, I chose a control period of one year prior to the Class Period. This is a common control period used in securities litigation cases as it is sufficiently long to account for seasonality, and it is outside the Class Period analyzed so it is not tainted by fraud-related events. After the fraud-related event days that could taint the control period are identified, a more targeted event analysis can be performed to more accurately quantify the Company-specific portion of a specific fraud-related price movement. In this case I used the 120 days prior to the event, excluding the price movements on the other fraud-related event days identified above. As a result, the control period is closer to the actual day analyzed, while still excluding fraud-related event days that could taint the analysis. A regression analysis using the price returns during the control period was then run to determine the regression equation used to predict Novatel's price returns on the days analyzed.
( Id. at ¶ 37.) Similarly, Defendants' expert Tabak acknowledges that a control period close to the event being analyzed should be used generally. (Tabak Dep., Doc. No. 323–6 at 94.) When asked about Steinholt's use of a 120–day control period in this situation, Tabak admitted that a 120–day control period would be reasonable, but voiced his concern that Steinholt had used a one-year control period initially and then adjusted the period to 120–days after learning the initial one-year control period results. ( Id. at 93–94.) Tabak's concerns reflect the crux of Defendants' objection to the 120–day control period. Defendants do not object to Steinholt's use of 120–day control period for any other reason than it was used after a one-year control period provided an allegedly unsatisfactory result. Defendants do not argue that Steinholt's findings when using 120–day control period are inaccurate or flawed. Nor do they suggest that using a one-year control period prior to a 120–day control period somehow influenced or changed the findings of the 120–day control period.
It should be noted that Steinholt utilized both a one-year and subsequently a 120–day control period to analyze the other relevant dates in this action, but July 20, 2007 is the only dated addressed by Defendants's motion. Thus, the Court limits its analysis to July 20, 2007, as well.
Having considered Steinholt's rationale for performing his analysis using a 120–day control period along with Defendants' objection to his doing so after performing his analysis using a one-year control period, the Court is not persuaded that Steinholt's use of the 120–day control period undermines the reliability of the event study. It seems unreasonable to limit an expert to the results of the first test performed when analyzing data simply because the first test had unfavorable results. Defendants have not established that the findings of the 120–day control period were unreasonable or unreliable based on the methodology of that test in particular. Defendants simply object to it being the second test performed. Had Steinholt simply utilized the 120–day control period initially, Defendants have proferred no evidence that the findings would have differed. Accordingly, Defendants have not persuaded the Court that Steinholt's use of the 120–day control period warrants exclusion of his report under Daubert.
Furthermore, Plaintiffs point out that Steinholt's loss causation does not rely upon the 120–day control period objected to by Defendants. Instead, Steinholt utilizes an intraday analysis of July 20, 2007, in order to establish loss causation. Accordingly, the Court will now consider Defendants objections to the admissibility of Steinholt's intraday analysis.
d. Steinholt's Use of Intraday Analysis to Establish Loss Causation on July 20, 2007
On July 20, 2007, Novatel stock initially declined in price and later that day the price began to increase. (Steinholt Rep., Doc. No. 299–4, ¶¶ 64, 67.) Rather than relying upon either the one-year or 120–day one-tailed tests to establish loss causation on July 20, 2007, Steinholt performed an intraday analysis to determine whether the price movements during July 20, 2007, were unusual. (Rebuttal ¶ 65.) In doing so, Steinholt compared the maximum intraday decline and rebound on July 20, 2007, to Novatel's historical maximum intraday declines and rebounds.
Having reviewed the parties' contentions regarding the intraday analysis and, more specifically, the parties' debate about the timing of the events occurring throughout the day, it appears that a brief summary of Steinholt's version of the events taking place around July 20, 2007, is appropriate. Prior to the market opening on July 20, 2007, a JMP Securities analyst report confirmed the cancellation of the U720 at Sprint (information that was not previously known to the market), provided new information about when the 727 would be available (also new information), and expressed the concern that Novatel would be losing market share to Sierra Wireless. (Steinholt Rep., Doc. No. 299–4, ¶ 62.) Following the issuance of the report, Novatel's stock price declined from a closing price of $28.37 per share on July 19, 2007, to a low of $25.09 on July 20, 2007, at 11:40 a.m. ( Id. at ¶ 64.) During this period of time, Defendants received several inquiries regarding the JMP report and Novatel's relationship with Sprint. ( Id. at ¶¶ 63–66.) Additionally, Defendant Weinert sent an email to a Craig–Hallum analyst regarding the JMP report in order to “reiterate that this ‘news' is no news to us, this has been expected, planned and forecasted for quite some time.” ( Id. at ¶ 65) Plaintiffs' contend that Craig–Hallum issued a report after receiving Weinert's email that stated, among other things, “ ‘NVTL shares were off sharply this morning, we believe in response to a rumor that NVTL may lose market share at Sprint.’ ” ( Id. at ¶ 66 (quoting Craig–Hallum Rep., July 20, 2007).) The Craig–Hallum report further stated that “ ‘[w]e understand this transition has been jointly planned for some time by NVTL and Sprint ... [and we] believe NVTL will continue to get the lion's share of Sprint's business.’ ” Id. Following the issuance of the Craig–Hallumreport, partly rebutting the JMP Securities report, Novatel's stock price increased from the $25.09 per share intraday low to a closing price of $27.05 per share. ( Id. at ¶ 67.) Based on this version of events, Steinholt's performed an intraday analysis to determine whether the price movements during July 20, 2007, were unusual.
Defendants raise several objections to Steinholt's intraday analysis, and the Court will address each objection individually. First, Defendants contend that the intraday analysis is inconsistent with the analysis Steinholt performed on the other relevant dates and that Steinholt “offers no explanation for why” he chose to change his methodology for this one date. (Defs.' Mot., Doc. No. 299–1 at 21.) As set forth above, there were several allegedly significant events that took place both before and throughout the day on July 20, 2007. It appears evident that Steinholt performed an intraday analysis in order to fully address the impact of each specific development. Here again, Defendants don't contend that his results are in error; they simply object to the “unexplained” departure from the original testing methodology. However, Steinholt did perform both the one-year control period analysis as well as the 120–day control period analysis with regard to this date. He simply chose to perform an additional test to more accurately reflect the nature of the events on that particular day. As noted above with regard to the 120–day control period, so long as the results are not in error and the test appears to be both relevant and reliable, the Court is disinclined to limit an expert to performing just one form of analysis.
Defendants' second objection to Steinholt's intraday analysis relates to Steinholt's conclusion that the Novatel's stock price declined 11.6% in response to the JMP report. (Defs.' Mot., Doc. No. 299–1 at 21.) Specifically, Defendants contend that Steinholt failed to address why “a substantial portion of the price decline occurred over an hour and a half after the market opened with full knowledge of the information disclosed in the JMP report” when reaching his conclusion. ( Id.) The Court finds this argument to be a thinly veiled attempt to contest Steinholt's conclusion whereas the reliability of Steinholt's methodology is more properly at issue in the Court's consideration of a Daubert motion.
Steinholt listed in his report all of the evidence he considered when determining that Novatel's stock price declined in response to the JMP report. Interpreting the stock price movement based on this type of evidence is well within Steinholt's expertise for Daubert purposes. Regardless, Defendants contend that without knowing the exact time of the Craig–Hallum Report there can be no way to determine whether Novatel's stock price reacted to the reports that day. The Court disagrees. Steinholt has not simply speculatedabout the timing of events on July 20, 2007; he has examined the available evidence in order to create an approximation upon which he can base his opinion. As Plaintiffs' point out, Defendants' expert also concluded that the decline on July 20, 2007 seemed to have been related to the discontinuance of the U720 disclosed in the JMP report. (Tabak Dep., Doc. No. 323–6 at 115.) Having reviewed the evidence that Steinholt considered with regard to the delay referenced by Defendants, the Court finds that Defendants' argument relates to the probative value of Steinholt's conclusion rather than its admissibility. Accordingly, to the extent that Defendants disagree with Steinholt's conclusion, they can challenge it during cross examination or offer the testimony of their own expert witness in this regard.
He considered the following:
(a) review of the JMP Securities analyst report; (b) the negative future cash flow implications of the discontinuation of the 720 USB modem by Sprint (a key customer); (c) internal e-mails showing the importance defendants placed on sales of Novatel's 720 USB modem to Sprint; (d) the large intra-day decline in Novatel stock price, statistically significant at the 99% confidence level; (e) the large reported trading volume in Novatel's stock, greater than 97% of the daily trading volumes on other days during the Class Period; (f) the contemporaneous Craig–Hallum analyst report attributing Novatel's stock price decline to the Sprint discontinuation of the 720 USB modem, and allegedly false commentary explaining the rebound in the stock price; and (g) internal Novatel e-mails attributing its stock price decline to the Sprint discontinuation of the 720 USB modem.
(Steinholt Rep., Doc. No. 299–4, ¶ 68.)
Defendants' third objection to Steinholt's intraday analysis pertains to probative value rather than admissibility for similar reasons. Defendants dispute Steinholt's conclusion that Novatel's stock price increased as a result of the Craig–Hallum report when Steinholt offers no evidence regarding when the Craig–Hallum report was released. Steinholt specifically addressed this issue in his rebuttal report wherein he identified the evidence used to approximate the timing of the Craig–Hallum report.
As I stated in my report, it is my understanding that the JMP Securities report was issued before the market open on July 20, 2007. Steinholt Report, ¶ 62. The earliest JMP Securities report time stamp I am aware of is 7:33:53AM. See Stoss Deposition, Exhibit 4. The earliest internal Novatel e-mail discussing the JMP Securities report is at 7:59AM (assuming West Coast time this translates into 10:59AM). (NOV–E–0579457). Discovery may provide more clarity on this issue, but the JMP Securities analyst report was definitely issued before the Craig–Hallum report. Regarding the Craig–Hallum report, the report itself provides critical information on when it was generated. Specifically, it references that “NVTL shares were off sharply this morning, we believe in response to a rumor that NVTL may lose market share at Sprint” (the topic of the JMP report), and provides the intra-day price of $27.12, which indicates that the earliest the report could have been generated is 10:59AM, the first time the intra-day price reached $27.12. Stoss Deposition, at 34–35. Furthermore, it is my understanding that the report was first publicly distributed at 11:35AM. See Stoss Deposition, Exhibit 17. Novatel's stock price reached its low a few minutes later at 11:40AM. Steinholt Report ¶ 64. The evidence I have reviewed is consistent with my opinions.
(Steinholt Rebuttal, Doc. No. 299–5, ¶ 68.) In contrast to Defendants' objection, it appears that Steinholt has considered and compiled the available evidence regarding the timing of the Craig–Hallum report and reached a conclusion on that basis. To the extent that Defendants disagree with Steinholt's assessment of the timing of the report, it becomes a question of fact for the jury to consider. Based on the evidence available to him, Steinholt reached a conclusion that appears to be logical and aligned with the stock price movements on July 20, 2007.
In sum, while Defendants may take issue with the details of his intraday methodology and his ultimate conclusions in that regard, the Court finds Defendants' arguments do not justify exclusion under Daubert, but will be more appropriately addressed at trial. As Steinholt states in his rebuttal report, the objective of his intraday analysis was to determine whether the price movements on July 20, 2007 were unusual. The intraday analysis appears to be justified by the unique factors of the case, particularly considering the stock price fluctuation occurring on July 20, 2007. This information could be helpful to a jury and Defendants' objections relate to the probative value of his analysis rather than its admissibility.
B. Steinholt's Damage Calculations
1. Steinholt's 10(b) Damages Analysis
When creating his event study, Steinholt first determined whether loss causation was established for class members who purchased prior to, and held or sold after, one of the four alleged corrective disclosure dates. After loss causation was established, Steinholt undertook the so-called out-of-pocket analysis in order to calculate Section 10(b) damages under the Securities Exchange Act of 1934. (Steinholt Rep., Doc. No. 229–4, ¶ 81.) “ ‘This test fixes recovery as the difference between the purchase price and the value of the security at the date of purchase less the difference between the sale price and the value of the security at the date of sale.’ ” ( Id. at ¶ 82 (quoting Bradford Cornell and R. Gregory Morgan, “ Using Finance Theory to Measure Damages in Fraud on the Market Cases,” 37 UCLA L. Rev., 883, 885 (June 1990)).) In his report, Steinholt further explains that “the most important part of calculating out-of-pocket damages is to determine the price at which the security would have traded absent the alleged fraud, also known as the value line.” ( Id. at ¶ 84 (emphasis added).) Having created the value line for the actual price of the stock without the influence of the alleged fraud, “the measure of out-of-pocket damages, as defined above, is either: (a) the artificial inflation per share at the time of purchase for shares purchased but not sold during the Class Period; or (b) the artificial inflation per share at the time of purchase less the artificial inflation per share at the time of sale for shares both purchased and sold during the Class Period.” ( Id. at ¶ 83.)
Those four dates being: July 20, 2007; February 20, 2007; April 14, 2008; and August 19, 2008.
Defendants argue that Steinholt's opinions on damages are unreliable and contrary to law after the Supreme Court's holding in Dura Pharmaceuticals, Inc. v. Broudo, 544 U.S. 336, 125 S.Ct. 1627, 161 L.Ed.2d 577 (2005). (Defs.' Mot., Doc. No. 299–1 at 22.) Specifically, Defendants contend that Steinholt employs an improper legal standard and premises his damage analysis on theories of loss causation rejected by the Supreme Court in Dura. ( Id.)
Defendants contend, as an initial matter, that Steinholt's damages analysis should be found unreliable because it is based upon an invalid event study and reference their previous arguments regarding Steinholt's loss causation methodology. As set forth above, the Court has determined that Steinholt's loss causation methodology survives Defendants' Daubert motion, and, thus, Defendants' objections to Steinholt's damages analysis on the basis of Steinholt's defective loss causation analysis are similarly unfounded.
In Dura, the Supreme Court considered whether economic loss and loss causation in a securities fraud case can be established simply by proving that the price on the date of purchase was inflated because of the misrepresentation. Dura, 544 U.S. at 342, 125 S.Ct. 1627. The Supreme Court found that price inflation alone was insufficient. Id. Following Dura, a plaintiff must demonstrate a causal connection between the material misrepresentation and the injury suffered by plaintiff. Dura, 544 U.S. at 342, 125 S.Ct. 1627;see also Metzler Inv. GMBH v. Corinthian Coll., Inc., 540 F.3d 1049, 1063 (9th Cir.2008) (citing In re Daou Systems, Inc., 411 F.3d 1006, 1025 (9th Cir.2005)).
Having reviewed the damage calculations utilized by Steinholt, the Court finds that his opinions are not contrary to the Supreme Court's holding in Dura. As an initial matter, Dura did not specifically address the proper damage calculations under Section 10(b). Dura simply requires plaintiffs to establish a causal connection between the alleged fraud and the injury in order to recover. While Dura does expressly warn against the use of price inflation alone to establish either economic loss or loss causation, Steinholt does not base his damage methodology solely on the price of inflation at the time of the stock purchase. Instead, his report reveals that Steinholt used an event study and regression analysis to determine the effect of the alleged misrepresentations and corrective disclosures on the inflated price of the stock. While inflated purchase price alone is not sufficient to show loss causation, the Court does not find Steinholt's damages methodology to be unreliable simply because it incorporates inflated purchase price into the overall calculation.
As discussed above, Steinholt addressed both industry and market factors when considering the impact of Defendants' alleged misrepresentations and the alleged corrective disclosures. Despite Defendants' contention that Steinholt's damages calculation fails to remove non-fraud-related factors from consideration, the calculation, in fact, depends heavily upon a value line representing the actual value of Novatel stock in the absence of the alleged fraud. (Steinholt Rep., Doc. No. 299–4, ¶ 84.) When creating the value line, he effectively replaced “the fraud-related price returns (relating both to the misrepresentations and the disclosures of the alleged truth) with the predicted price returns, while holding all other price returns constant.” ( Id. at ¶ 85.) Additionally, Steinholt excluded damages for shares purchased and sold during periods where loss causation had not been established. ( Id.) Having reviewed Steinholt's process, it is evident that Steinholt relies upon much more than simply price inflation when calculating damages under Section 10(b). Accordingly, Defendants' Dura argument is unavailing.
Similarly, the Court is unpersuaded by Defendants reliance upon In re Williams Securities Litigation, 496 F.Supp.2d 1195 (N.D.Okla.2007), aff'd, 558 F.3d 1130 (10th Cir.2009). In Williams, the expert did not identify the dates of the alleged corrective disclosures nor did he distinguish between fraud-related and non-fraud-related causes for stock prices decline, i.e. industry factors, other company news, and other negative developments unrelated to the alleged fraud. The court found that the expert's event study was deficient for all of these reasons and, thus, his damages calculations were also deemed unreliable for failure to consider all of these factors. Here, in contrast to the situation in Williams, the Court has found Steinholt's event study to be reliable on the whole and Steinholt's damages analysis utilizes a value line taking into account the value of the stock absent the alleged fraud. While it does appear that the court in Williams interpreted Dura as precluding the type of damages analysis performed in this instance by Steinholt, the Court declines to follow this interpretation for the reasons set forth above.
To the extent that Defendants suggest that there is a possibility that plaintiffs could recover for losses that are non-fraud-related, there is no evidence that this risk undermines the reliability of Steinholt's methodology as a whole. Steinholt has attempted to minimize the risk of recovering damages for non-fraud-related losses. Steinholt's analysis has limited to recovery for only those parties for whom loss causation has been established, and he has removed non-fraud-related industry and market factors from his calculations. Accordingly, the Court finds that it is not an issue of admissibility, but rather of probative value to be addressed at trial. Based on the foregoing, the Court denies Defendants' motion to exclude Steinholt's calculations of Section 10(b) damages.
2. Steinholt's Section 20(a) Damages Analysis
Damages under Section 20A of the Exchange Act are limited to “profit gained or loss avoided.” 15 U.S.C. § 78t–1(b)(1). Based upon this statutory language, Steinholt expressed his understanding that “Section 20A damages are either calculated based on: (a) the inflation per share that existed at the time of the insider sales, or (b) the losses avoided by the defendants by selling the shares during the Class Period, as opposed to selling the shares following the disclosure of the relevant truth at the end of the Class Period.” (Doc. 299–4 at ¶ 90.) Steinholt determined that the profits gained were less than the losses avoided in all instances and, for this reason, Steinholt utilized the profits gained measurement to determine Defendants Section 20A damages. ( Id.)
Unsurprisingly, Defendants raise multiple objections to Steinholt's opinion in this regard. First, Defendants contend that Steinholt's Section 20A analysis fails to comport with Dura. The Court dispenses with this argument for the same reasons set forth above with regard to Section 10(b) damages. Dura does not articulate a standard for calculating damages, and the Court will not extend its holding in order to do so here.
Second, Defendants contend that Steinholt's analysis leads to the “absurd result that each defendant's ‘profits' are greater than their total sales proceeds less the amount each defendant paid for his or her shares.” (Defs.’ Mot., Doc. No. 299–1 at 25.) Specifically, Defendants object to the possibility that the damages found by Steinholt could be greater than the difference between what the defendant paid for the stock initially and what the defendant sold the stock for prior to the corrective disclosures. While the Court recognizes the logic within Defendants' theory, the same can be said for Steinholt's method. Defendants have not proffered any controlling legal authority for finding Steinholt's method unreliable under Daubert, nor has the Court found any in its own research on the issue. Faced with two differing, but reasonable, theories on how to calculate profits gained, the Court finds that the issue is not one of reliability to be decided under Daubert. Having reviewed Steinholt's method of calculation and finding it well-reasoned, the Court finds that Defendants' objection to Steinholt's results pertain to weight rather than admissibility.
Here is the example provided by Defendants to demonstrate the allegedly “absurd result” reached by Steinholt's calculations. Defendant George B. Weinert purchased 3,500 shares for $18.78 per share and sold the same shares for $24.00. Defendants contend that his profit could be, at most, the difference between the $18.78 purchase price and the $24.00 sale price, meaning $5.22 per share. In contrast, Steinholt determined that Weinert had received a “profit” of $12.47 because the true value of the share at the time Weinert sold the share for $24.00 was only $11.53.
The only legal authority proffered by Defendants for finding Steinholt's profit gained methodology inadmissible is neither persuasive nor controlling in this instance. Short v. Belleville Shoe Mfg., Co., 908 F.2d 1385, 1392 (7th Cir.1990). Id. The standard that Defendants cite for how Section 20A “damages are ordinarily measured” is set forth in the midst of a long discussion regarding equitable tolling and statutes of limitations for securities claims. The court simply stated that “[p]rofit gained or loss avoided is a contemporaneous measure: the difference between the price the insider realizes and the market price of the securities after the news is released.” Id. There was no further discussion regarding the proper method of calculation. This excerpt, while providing support for the methodology offered by Defendants, does not impeach that offered by Steinholt. It is often the case that competing methodologies exist for reaching a certain result, and Short is not a sufficient basis for finding Steinholt's to be unreliable in this instance.
Third, Defendants contend that Steinholt provides no explanation in his report for why he excluded Hadley and Souissi sales prior to May 18, 2007 from his Section 20A damages calculations. (Defs.' Mot., Doc. 299–1 at 25.) When asked about the exclusion at his deposition, Steinholt indicated that he had excluded the sales on Plaintiffs' instruction. (Steinholt Dep., Doc. 299–9 at 48.) Defendants suggest that Steinholt will be unable to testify reliably about his overall methodology because he cannot explain his rationale for excluded the Hadley and Souissi sales prior to May 18, 2007. Once again, the Court finds that Defendants' argument pertains to weight rather than admissibility. The methodology used by Steinholt remains the same, but with different parameters. Defendants do not suggest that the methodology itself is in error or unreliable. Accordingly, Defendants may cross-examine Steinholt at trial regarding the exclusion of Hadley and Souissi sales prior to May 18, 2007 and his reasons for doing so, but the calculations as a whole will not be excluded on this basis.
For the reasons set forth above, the Court finds Steinholt's methodology regarding Section 20A damages to be reliable and relevant under the Daubert standard.
3. Steinholt's Damages Analysis Regarding March 26, 2007
In his report, Steinholt opines that Defendants' “allegedly false statements relating to its 1Q07 guidance caused Novatel's stock price to increase on March 26, 2007, and to trade at artificially inflated prices.” (Doc. 299–4, ¶¶ 50–52.) Defendants object to this analysis being included in Steinholt's report because Plaintiffs have not alleged in their complaint that Defendants made false statements on March 26, 2007. Accordingly, Defendants seek to exclude Steinholt's damages analysis regarding March 26, 2007, and the entirety of his damages analysis to the extent that it is impacted by Steinholt's inclusion of March 26, 2007, in his overall damages calculation.
After reviewing the record with regard to the fraud allegations associated with March 26, 2007, the Court finds Defendants' objection to be unwarranted. Plaintiffs' original complaint expressly referenced the Novatel press release issued on March 26, 2007, and analyzed by Steinholt in his report. (Doc. 1, p. 10.) Following the Court's order consolidating the case, Plaintiffs' filed a reformatted, amended complaint. (FAC, Doc. No. 23.) While the amended complaint does not retain the specific reference to the March 26, 2007 press release from the initial complaint, the amended complaint alleges that “Novatel's financial results concerning revenues and earnings, reported in press releases, SEC filings and conference calls, were materially misleading and did not fairly present the financial condition of the Company through fiscal quarter ended March 31, 2007.” (Doc. 23, p. 21.) This statement encompasses the alleged fraud contained within the March 26, 2007 press release. For this reason, the Court finds Defendants suggestion that the fraud allegations with regard to the March 26, 2007, press release constitute “new facts” to be somewhat disingenuous.
Furthermore, although Plaintiffs' amended complaint does not specifically identify the March 26, 2007, press release, Steinholt's report certainly does. (Doc. 299–4, p. 31). Steinholt executed his report on December 13, 2010, giving Defendants notice of his opinion two months prior to the filing of Defendants' motion for summary judgment. Not only did Steinholt fully address the March 26, 2007, press release; Defendants' own expert spent considerable time in his report criticizing him for doing so. (Doc. 299–10, p. 21–23.) Steinholt then discussed the fraud allegations regarding the March 26, 2007, press release in his deposition, taken February 1, 2011, (Doc. No. 299–9, p. 4), and Plaintiffs' referenced the allegations in their response to Defendant Hadley's interrogatories on February 11, 2011. (Doc. 322–4, p. 94). Based on this time line, Defendants had notice of Plaintiffs' March 26, 2007, allegations at least two months prior to filing their motion for summary judgment regardless of its being contained within Plaintiffs' complaints.
To the extent that Plaintiff's March 26, 2007, fraud allegations are not found to be encompassed by allegations within the amended complaint, the Court finds it appropriate to allow Plaintiff to amend it accordingly. Federal Rule of Civil Procedure 15(a) provides that leave of the court allowing a party to amend its pleading “shall be freely given when justice so requires.” Leave to amend lies within the sound discretion of the trial court, which discretion “must be guided by the underlying purpose of Rule 15 to facilitate decision on the merits, rather than on the pleadings or technicalities.” United States v. Webb, 655 F.2d 977, 979 (9th Cir.1981) (citations omitted). Thus, Rule 15's policy of favoring amendments to pleadings should be applied with “extreme liberality.” Id.; DCD Programs, Ltd. v. Leighton, 833 F.2d 183, 186 (9th Cir.1987) (citations omitted).
Here, Defendants have not alleged that allowing Plaintiffs to amend their complaint would create undue delay, bad faith or dilatory motive, futility of amendment, and prejudice to the opposing party. See Foman v. Davis, 371 U.S. 178, 182, 83 S.Ct. 227, 9 L.Ed.2d 222 (1962). Nor does the Court find that any such reason exists. It is apparent that Defendants' expert had the opportunity to address the March 26, 2007, fraud allegations in his report and Defendants had the opportunity to do so in their motion for summary judgment. Therefore, the portion of Plaintiffs' response to Defendant Hadley's Interrogatory No. 21 relevant to March 26, 2007, is deemed part of the complaint. Defendants are not required to answer.
The Court includes the relevant language, omitting a lengthy quote from the actual press release, here:
Novatel's financial guidance for 1Q07 was false and/or misleading. On March 26,2007, Novatel provided new financial guidance for the Company's 1Q07 revenues and earnings significantly above its previous guidance. The Company's press release stated:
[Press release quotation omitted.]
These statements were false and/or misleading because Novatel's financial guidance for 1Q07 financial results were false and/or misleading. Novatel's cumulative revenue recognition violations, its undisclosed channel stuffing practices and undisclosed earnings trends were material to each of its quarters during the Class Period under the provisions of SAB No. 99. All accounting violations have been summarized by quarter at Exhibits C(1)-C(3) of the Regan Report. Also, the Company failed to disclose a material weakness within its internal controls over financial reporting during 1Q07 as required under relevant SEC Rules and the Sarbanes–Oxley Act of 2002. See Regan Report.
(Steinholt Rep., Doc. No. 322–4 at 97–98.)
Insomuch as Plaintiffs' complaint has been amended to include the fraud allegations regarding March 26, 2007, Defendants' objection to Steinholt's damages analysis on the basis of his including allegations not asserted in the complaint is now moot. For all of the reasons set forth above, Steinholt's damage calculations are deemed admissible.
Conclusion
For the reasons set forth above, the Court rules as follows: (1) Defendants' motion to exclude the opinion and testimony of Bjorn Steinholt, Doc. No. 299, is DENIED, and (2) Plaintiffs' amended complaint is amended to include the March 26, 2007, fraud allegations discussed above.
IT IS SO ORDERED.