Smilovits v. First Solar Inc.

Full title: Mark Smilovits, et al., Plaintiffs, v. First Solar Incorporated, et al.…

Court: UNITED STATES DISTRICT COURT FOR THE DISTRICT OF ARIZONA

Date published: Aug 10, 2015

119 F. Supp. 3d 978 (D. Ariz. 2015)

Opinion

No. CV-12-00555-PHX-DGC

08-10-2015

Mark Smilovits, et al., Plaintiffs, v. First Solar Incorporated, et al., Defendants.

Garrett Webster Wotkyns, Schneider Wallace Cottrell Konecky LLP, Scottsdale, AZ, Jennifer Lynn Kroll, Susan Joan Martin, Martin & Bonnett PLLC, Andrew S. Friedman, Kevin Richard Hanger, Bonnett Fairbourn Friedman & Balint PC, Phoenix, AZ, Jeremy A. Lieberman, Pomerantz LLP, Richard W. Gonnello, Faruqi & Faruqi LLP, New York, NY, Patrick V. Dahlstrom, Pomerantz LLP, Chicago, IL, Daniel S. Drosman, Danielle S. Myers, Jason A. Forge, Luke Brooks, Mark Solomon, Michael J. Dowd, Christopher Dennis Stewart, Cody R. Lejeune, Darren J. Robbins, Darryl J. Alvarado, Jennifer N. Caringal, Lonnie A. Browne, Tor Gronborg, Robbins Geller Rudman & Dowd LLP, San Diego, CA, for Plaintiffs. Anna Erickson White, Jordan Eth, Eugene G. Illovsky, James P. Bennett, Judson E. Lobdell, Mark Ryan Scott Foster, Paul Flum, Philip T. Besirof, Morrison & Foerster LLP, San Francisco, CA, Joseph Nathaniel Roth, Maureen Beyers, Osborn Maledon PA, Phoenix, AZ, for Defendants.

David G. Campbell, United States District Judge

Garrett Webster Wotkyns, Schneider Wallace Cottrell Konecky LLP, Scottsdale, AZ, Jennifer Lynn Kroll, Susan Joan Martin, Martin & Bonnett PLLC, Andrew S. Friedman, Kevin Richard Hanger, Bonnett Fairbourn Friedman & Balint PC, Phoenix, AZ, Jeremy A. Lieberman, Pomerantz LLP, Richard W. Gonnello, Faruqi & Faruqi LLP, New York, NY, Patrick V. Dahlstrom, Pomerantz LLP, Chicago, IL, Daniel S. Drosman, Danielle S. Myers, Jason A. Forge, Luke Brooks, Mark Solomon, Michael J. Dowd, Christopher Dennis Stewart, Cody R. Lejeune, Darren J. Robbins, Darryl J. Alvarado, Jennifer N. Caringal, Lonnie A. Browne, Tor Gronborg, Robbins Geller Rudman & Dowd LLP, San Diego, CA, for Plaintiffs.

Anna Erickson White, Jordan Eth, Eugene G. Illovsky, James P. Bennett, Judson E. Lobdell, Mark Ryan Scott Foster, Paul Flum, Philip T. Besirof, Morrison & Foerster LLP, San Francisco, CA, Joseph Nathaniel Roth, Maureen Beyers, Osborn Maledon PA, Phoenix, AZ, for Defendants.

ORDER

David G. Campbell, United States District Judge

In this complex securities fraud class action, Defendants have filed a motion for summary judgment on all claims (Doc. 311) and Plaintiffs have filed a motion for partial summary judgment on eighteen affirmative defenses (Doc. 309). Defendants have also filed a request for judicial notice (Doc. 341) and two motions to seal (Docs. 342, 387). Each motion has been briefed, and the Court heard oral argument on July 22, 2015. The Court will deny in part and grant in part Defendants' motion for summary judgment, deny Plaintiffs' motion for summary judgment, and grant Defendants' request for judicial notice and motions to seal.

The Court finds two competing lines of cases in the Ninth Circuit on loss causation. Because one line would result in complete summary judgment for Defendants and the other (which the Court chooses to follow) will result largely in denial of summary judgment and a lengthy and expensive trial, the Court will certify this issue for immediate appeal under 28 U.S.C. § 1292(b).

I. Background.

First Solar, Inc. is one of the world's largest producers of photovoltaic solar panel modules. Its stock is publicly traded on the NASDAQ Global Market. By 2008, First Solar's stock had risen to nearly $300 per share. As of the beginning of 2012, the stock price had fallen to less than $50 per share. During this time, which coincided with the recession in 2008, First Solar experienced a change in leadership, a manufacturing defect, and a climate-related technical issue regarding their modules.

Plaintiffs are purchasers of First Solar stock who brought this class action alleging that First Solar and several of its key officers and executives misrepresented the financial state of the company to inflate the price of First Solar stock, committed accounting violations, and concealed material facts relating to the extent of the manufacturing defect and the hot climate issue in violation of §§ 10(b) and 20(a) of the Securities Exchange Act of 1934 and Securities Exchange Commission Rule 10b-5. Plaintiffs filed their First Amended Complaint on August 17, 2012, and the Court certified Plaintiffs' class on October 8, 2013. Fact discovery has been completed. Expert discovery remains.

A. The Parties.

The class is defined as “[a]ll persons who purchased or otherwise acquired the publicly traded securities of First Solar, Inc. between April 30, 2008 and February 28, 2012” (the “Class Period”). Doc. 171 at 22.¹

1 Page citations to electronically filed documents refer to the stamped CM/ECF page numbers at the top of each page, not the original document's page numbers.

First Solar, Inc. is headquartered in Tempe, Arizona. During the Class Period, it operated manufacturing facilities in Ohio, Germany, and Malaysia. First Solar is managed by a shareholder-elected Board of Directors. The Board delegates functions to committees within the company, including the Audit Committee, which performs internal accounting audits. First Solar's accounting practices are also audited and reviewed by PricewaterhouseCoopers (“PwC”), an outside accounting firm.

The Individual Defendants consist of several officers and executives employed by First Solar. Michael Ahearn was the Executive Chairman of the Board throughout the entire Class Period. Doc. 312; Doc. 363 at 13. He also served as the Chief Executive Officer (“CEO”) from April 2008 to October 2009 and from October 2011 to the end of the Class Period. Doc. 363 at 13. Robert Gillette served as CEO and Director of First Solar from October 2009 to October 2011. Id. Bruce Sohn served as President from the beginning of the Class Period until April 2011. Id. at 14. David Eaglesham served as Vice President (“VP”) of Technology from the beginning of the Class Period until November 2009, when he became Chief Technology Officer. Doc. 312. Jens Meyerhoff served as Chief Financial Officer (“CFO”) from the beginning of the Class Period until December 2010, and then assumed the role of President of the Utility Systems Business Group. Id. James Zhu served as VP and Corporate Controller, then VP and Chief Accounting Officer, and finally as the Interim CFO. Id. Mark Widmar took over Zhu's role as CFO in April 2011. Id.

Several other individuals employed during the Class Period, but not named as defendants in this action, performed key roles. These include Michael Koralewski, who served separately as Director of Global Quality, then as VP of Global Quality, and later as VP of Site Operations and Plant Manager; TK Kallenbach, who served separately as Executive VP of Marketing and Product Management and later as President of the Components Business Group; Thomas Kuster, who served briefly as VP of Engineering Procurement and Construction and then as VP of System Development; and Bryan Schumaker, who served as Assistant Corporate Controller and later as VP and Corporate Controller. Doc. 312.

B. The LPM Defect.

In March 2009, First Solar received a complaint from one of its German customers that some of its sites were experiencing low power output. Doc. 332 at 19.² A few months later, a task force led by Eaglesham discovered that the power loss was the result of a new manufacturing process implemented in June 2008. Doc. 314, ¶¶ 10-11. The process “had the effect of producing a small subpopulation of modules that could experience field power loss of 15% or more from nameplate within the first several months of installation.” Id., ¶ 12. The modules became known as Low Power Modules (“LPMs”), and the defective manufacturing process was discontinued in June 2009. Id., ¶ 14.³

2 First Solar tested solar panel modules as they came off the assembly line. Doc. 363 at 15. This “destructive testing [was done] to simulate performance following installation in the field,” and the results were referred to as a Stability Index (“STBi”). Doc. 311 at 28 n.12. The STBi data was the key metric used to “estimate the number of modules that could experience premature power degradation.” Doc. 363 at 15.

3 First Solar warranted that their modules would “produce at least (1) 90% of their labeled power during the first ten years after their sale and (2) 80% of their labeled power during years eleven to twenty-five.” Doc. 311 at 17; Doc. 334 at 12. The warranty required the customer to ship a defective module to First Solar, where it would be tested to confirm underperformance. First Solar would ship a new module to the customer. Doc. 334 at 12. Expected costs from warranty claims were estimated by First Solar and included as a Warranty Accrual line item in its financial reports. Doc. 311 at 19.

Shortly after discovering the defect, First Solar agreed to remediate sites affected by LPMs. It contacted customers to notify them of the defect and offered remediation by removing and replacing LPMs at sites that were underperforming. Customers were required to submit remediation claims by November 2010. Doc. 324, ¶ 20.

In order to account for the added expense of remediation in First Solar's financial statements, Koralewski developed models for estimating the number of LPMs that were produced between June 2008 and June 2009. Id., ¶¶ 9-12. At the time, he believed First Solar “could identify LPMs by serial numbers and replace only those modules.” Id., ¶ 21. After it became clear that First Solar could not merely replace single LPMs, Koralewski was again charged with estimating the number of modules required to remediate customer sites. Id. These estimates were based on various statistical models and accounted for “hit rate calculations,” which “refer to the percentage of returned modules that were LPMs.” Id., ¶ 22b. For example, “[f]or small rooftop sites, which usually contained hundreds of modules, [First Solar] determined that it was more efficient to replace all of the modules rather than search for LPMs individually.” Id. This required First Solar to replace a greater number of modules than initially anticipated.

In the quarters immediately following discovery of the LPM defect, Koralewski reported his estimates internally to First Solar executives. In the third quarter of 2009 (“3Q09”), Koralewski estimated that there were 115,000 LPMs in the field. Id., ¶ 11. In 4Q09, the estimate grew to 154,000. Id. By 1Q10, Koralewski estimated that 450,000 modules were LPMs, which represented less than 4% of the total 11.8 million modules produced during the defect period. Id., ¶ 12.

The estimates regarding the number of LPMs in the field and the number of modules required to remediate the defect directly affected the additional costs First Solar faced as a result of the manufacturing defect. The costs were reflected in the “LPM Remediation Accrual,” which was calculated to account for the additional expenses in accordance with Generally Accepted Accounting Principles (“GAAP”). Doc. 325, ¶ 24. Over the course of several quarters, the LPM Remediation Accrual grew with the estimated number of modules required to complete remediation.

Another factor that contributed to the estimate was the number of customer claims First Solar received, as well as the percentage of those claims that First Solar believed valid. After initially contacting customers, First Solar had completed remediation of “more than two dozen of the approximately 150 sites that had been claimed[.]” Doc. 324, ¶ 25. But in the weeks leading up to the November 2010 deadline, the company “received over 5,000 new claims, most of which were not accompanied by supporting data.” Id., ¶ 26.

The LPM manufacturing defect and the resulting remediation costs were not disclosed to the public until July 2010, when the LPM Remediation Accrual appeared as a separate line-item in First Solar's 2Q10 Form 10-Q accompanied by the following explanation:

During the period from June 2008 to June 2009, a manufacturing excursion occurred affecting less than 4% of the total product manufactured within the period. The excursion could result in possible premature power loss in the affected modules. The root cause was identified and subsequently mitigated in June 2009. On-going testing confirms the corrective actions are effective. We have been working directly with impacted customers to replace the affected modules and these efforts are well underway and, in some cases, complete. Some of these efforts go beyond our normal warranty coverage. Accordingly, we have accrued additional expenses of $17.8 million in the second quarter of 2010 and $29.5 million in total to date to cover the replacement of the anticipated affected module population in the field.

Doc. 359-1 at 41.⁴

4 Each quarter, First Solar issued Forms 10-Q or 10-K depending on whether the report pertained to the first three quarters (10-Q) or the full year (10-K). First Solar also participated in earnings calls with securities analysts when the 10-Qs and 10-Ks were released, and the calls were open to the public. Doc. 311 at. 20.

In 3Q10, the figures remained the same. Doc. 325, ¶ 37. In 4Q10, the LPM Remediation Accrual grew by $8.5 million. Doc. 331 at 118. In 1Q11, the figures did not increase, and in 2Q11, the figures increased by $3.6 million. Doc. 325, ¶¶ 39, 41. In 3Q11, $22.1 million was added to the LPM Remediation Accrual. By 4Q11, 90% of the outstanding claims had been processed, and the figures were increased by $23.9 million with a $70.1 million product warranty expense. Doc. 340 at 6-7.

Plaintiffs argue that First Solar wrongfully failed to disclose the LPM defect prior to July 29, 2010. Plaintiffs further assert that First Solar misrepresented the true scope of the defect by engaging in improper accounting practices and reporting false information on their financial statements.

C. Hot Climate Degradation.

In April 2010, a team of First Solar scientists discovered data suggesting that First Solar modules installed in hot climates experienced faster power loss than previously understood. Doc. 314, ¶ 32. This data, however, was inconsistent with recent data indicating that “long-term test installations in the Arizona desert” were performing above expectation ratios. Id., ¶¶ 33(b), (c). The team continued to monitor sites.

On February 7, 2011, First Solar discovered that the company's Blythe, California plant was producing power at a lower level than its Ontario, Canada plant. Id., ¶¶ 35-36. In March, the team of scientists concluded that the modules were experiencing a greater “initial stabilization” in hot climates than previously understood. Id., ¶ 38. Mitigation strategies were implemented, and Koralewski concluded that First Solar's existing warranty accrual was sufficient to cover projected warranty claims from customers. At the end of 4Q11, the hot climate degradation had been resolved, and the Warranty Accrual line item was increased by $37.8 million.⁵

5 First Solar ultimately determined that the hot climate degradation affected almost 10 million modules produced between July 2009 and June 2011. Doc. 364-2 at 7.

Plaintiffs argue that First Solar wrongfully concealed the hot climate defect for several quarters by manipulating accounting metrics and ignoring the true scope of the defect. They also allege that First Solar buried the extra costs of the hot climate defect in its Warranty Accrual instead of disclosing it in a separate line item.⁶

6 Plaintiffs also argue that First Solar manipulated one of its “key metrics”—cost-per-watt (“CpW”). CpW is defined as “the total manufacturing cost incurred during a period divided by the total watts produced during that period.” Doc. 363 at 53. They assert that VP Kurt Woods pressured employees to “bring the cost per watt down” one cent, which was reported to the Internal Audit Committee (“IA”), and an investigation was undertaken. Plaintiffs assert that the investigation was cut short and Meyerhoff ordered that no employee should report improper conduct to IA again. Plaintiffs also assert that First Solar engaged in improper accounting methods to manipulate CpW.

D. The Trades.

During the Class Period, the Individual Defendants made several trades of First Solar stock. Ahearn sold over three million shares in multiple trades, amounting to more than 96% of his shares. Doc. 363 at 55. Eaglesham sold 94% of his stock over several trades, and Meyerhoff sold over 80% of his shares. Id. at 57-58. Sohn sold nearly 75% of his shares, and Zhu sold nearly 50%. Id. at 58. In contrast, both Gillette and Widmar purchased several thousand shares of First Solar stock. Id. at 57-58. Plaintiffs assert that the timing of the sales shows that Defendants knew the LPM defect was going to cost much more than First Solar had reported in its financial statements.

E. Value of First Solar's Stock.

First Solar stock experienced several days of steep declines during the Class Period, which appeared to be market reactions to quarterly financial disclosures and the departure of Gillette as CEO. On July 29, 2010, First Solar announced its 2Q10 earnings, which disclosed the manufacturing defect and additional costs of $23.4 million. Id. at 65-66. The stock price dropped 7.4% the next day. Id. at 66. On February 24, 2011, First Solar announced its 4Q10 earnings, missing its target revenue. Id. at 67. The stock price declined by 5.4% the next day. Id. at 68. On May 3, 2011, the company announced its 1Q11 earnings, which included additional expenses for LPM remediation. Id. at 69. The next day, First Solar stock dropped 6.2%. Id. On October 25, 2011, First Solar announced Gillette's departure as CEO. Id. at 70. The stock price dropped 25% that day, but later rebounded. Id. On December 14, 2011, the company issued a press release and held a conference call relating to its financial state. Id. at 73. First Solar stock dropped an additional 21.4%. Id. On February 28, 2012, First Solar announced disappointing 4Q11 results. Id. at 74-75. The stock price dropped 11.26% that day and 5.8% the next. Id. at 75.

II. Legal Standard.

A party seeking summary judgment “bears the initial responsibility of informing the district court of the basis for its motion, and identifying those portions of [the record] which it believes demonstrate the absence of a genuine issue of material fact.” Celotex Corp. v. Catrett, 477 U.S. 317, 323, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986). Summary judgment is appropriate if the evidence, viewed in the light most favorable to the nonmoving party, shows “that there is no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law.” Fed. R. Civ. P. 56(a). Summary judgment is also appropriate against a party who “fails to make a showing sufficient to establish the existence of an element essential to that party's case, and on which that party will bear the burden of proof at trial.” Celotex, 477 U.S. at 322, 106 S.Ct. 2548. Only disputes over facts that might affect the outcome of the suit will preclude the entry of summary judgment, and the disputed evidence must be “such that a reasonable jury could return a verdict for the nonmoving party.” Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 248, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986).

Plaintiffs allege that Defendants violated § 10(b) of the Securities Exchange Act of 1934 and Securities Exchange Commission Rule 10b-5. Section 10(b) “makes it unlawful to ‘use or employ, in connection with the purchase or sale of any security...any manipulative or deceptive device or contrivance in contravention of such rules and regulations as the Commission may prescribe.’ ” In re Oracle Corp. Sec. Litig., 627 F.3d 376, 387 (9th Cir.2010) (quoting 15 U.S.C. § 78j(b)). “Commission Rule 10b-5 forbids, among other things, the making of any ‘untrue statement of a material fact’ or the omission of any material fact ‘necessary in order to make the statements made...not misleading.’ ” Dura Pharm., Inc. v. Broudo, 544 U.S. 336, 341, 125 S.Ct. 1627, 161 L.Ed.2d 577 (2005) (quoting 17 C.F.R. § 240.10b-5 (2004)). “The scope of Rule 10b-5 is coextensive with that of Section 10(b).” Oracle, 627 F.3d at 387. To demonstrate a violation of § 10(b) and Rule 10b-5, “a plaintiff must prove (1) a material misrepresentation or omission by the defendant; (2) scienter; (3) a connection between the misrepresentation or omission and the purchase or sale of a security; (4) reliance upon the misrepresentation or omission; (5) economic loss; and (6) loss causation.” Stoneridge Inv. Partners, LLC v. Scientific – Atlanta, Inc., 552 U.S. 148, 157, 128 S.Ct. 761, 169 L.Ed.2d 627 (2008). Defendants' motion for summary judgment asserts that Plaintiffs cannot prove elements (1), (2), and (6), but focuses first and most extensively on loss causation.

III. Loss Causation, Ninth Circuit Law, and § 1292(b) Certification.

Plaintiffs assert that loss causation is satisfied if the facts misrepresented or omitted by Defendants ultimately cause Plaintiffs' loss. Under their view, “ ‘a plaintiff can satisfy loss causation by showing that the defendant misrepresented or omitted the very facts that were a substantial factor in causing the plaintiff's economic loss.’ ” Doc. 363 at 62 (quoting Nuveen Mun. High Income Opportunity Fund v. City of Alameda, 730 F.3d 1111, 1120 (9th Cir.2013) (emphasis in Nuveen; citation in Nuveen omitted)). Defendants favor a narrower definition. They argue that loss causation can be established only if “ ‘the market learns of a defendant's fraudulent act or practice, the market reacts to the fraudulent act or practice, and plaintiff suffers a loss as a result of the market's reaction.’ ” Doc. 379 at 13 (quoting Oracle, 627 F.3d at 392). Each side cites Ninth Circuit cases in support of its position. The Court has read the Ninth Circuit cases cited by the parties—several times—and concludes that they reflect two irreconcilable lines of cases. The Court will provide a brief history of loss causation, describe each line of Ninth Circuit cases, and then decide which line to follow.

A. A Brief History of Loss Causation.

As far back as the early 1980s, some federal courts recognized that a securities fraud plaintiff should be permitted to recover under § 10(b) and Rule 10b-5 only if the misrepresentation of omission of the defendant proximately caused the plaintiff's loss. A leading case was Huddleston v. Herman & MacLean, 640 F.2d 534 (5th Cir.1981), aff'd in part, rev'd in part, 459 U.S. 375, 103 S.Ct. 683, 74 L.Ed.2d 548 (1983), which held that “[t]he plaintiff must prove not only that, had he known the truth, he would not have [purchased the security], but in addition that the untruth was in some reasonably direct, or proximate, way responsible for his loss,” id. at 549. “If the investment decision is induced by misstatements or omissions that are material and that were relied on by the claimant, but are not the proximate reason for his pecuniary loss, recovery under [Rule 10b-5] is not permitted.” Id. Without this requirement of proximate cause, Huddleston explained, “Rule 10b-5 would become an insurance plan for the cost of every security purchased in reliance upon a material misstatement or omission.” Id.

The Huddleston view was not universally accepted. Some courts held that a plaintiff could prevail merely by showing that the misrepresentation or omission caused the plaintiff to purchase the security. See, e.g., Kafton v. Baptist Park Nursing Ctr., Inc., 617 F.Supp. 349, 350 (D.Ariz.1985). This broader form of causation is sometimes called “transaction causation.” It exists when a misrepresentation or omission of the defendant induces the plaintiff to purchase the defendant's securities. As the Ninth Circuit has explained, “to prove transaction causation, the plaintiff must show that, but for the fraud, the plaintiff would not have engaged in the transaction at issue.” In re Daou Systems, Inc., 411 F.3d 1006, 1025 (9th Cir.2005). “[T]o prove loss causation, the plaintiff must demonstrate a causal connection between the deceptive acts that form the basis for the claim of securities fraud and the injury suffered by the plaintiff.” Id.

A helpful illustration of the difference was provided in Huddleston:

[A]n investor might purchase stock in a shipping venture involving a single vessel in reliance on a misrepresentation that the vessel had a certain capacity when in fact it had less capacity than was represented in the prospectus. However, the prospectus does disclose truthfully that the vessel will not be insured. One week after the investment the vessel sinks as a result of a casualty and the stock becomes worthless.

640 F.2d at 549 n. 25. In this example, the investor might be able to prove transaction causation (that the misrepresentation about the vessel's capacity induced him or her to purchase the stock), but could not prove loss causation (that the misrepresentation caused the investor's loss). The loss was caused by the lack of insurance.

Although federal courts disagreed for several years on whether loss causation was required in 10b-5 cases, Congress resolved the disagreement in 1995 when it passed the Private Securities Litigation Reform Act (“PSLRA”). The PSLRA required proof of transaction causation by requiring proof of reliance—that the plaintiffs relied on the defendant's misstatement or omission when they purchased the security. See Dura, 544 U.S. at 341, 125 S.Ct. 1627 (a 10b-5 plaintiff must prove “reliance, often referred to in cases involving public securities markets (fraud-on-the-market cases) as ‘transaction causation’ ”). The PSLRA also included a section titled “Loss causation” which provided that “[i]n any private action arising under this chapter, the plaintiff shall have the burden of proving that the act or omission of the defendant alleged to violate this chapter caused the loss for which the plaintiff seeks to recover damages.” 15 U.S.C. § 78u-4(b)(4). As the Supreme Court has noted, this provision requires proof of “ ‘loss causation,’ i.e., a causal connection between the material misrepresentation and the loss[.]” Dura, 544 U.S. at 341, 125 S.Ct. 1627. Loss causation has thus become a universal requirement of securities fraud cases.

The Supreme Court addressed the requirement of loss causation in Dura. Some courts had held that loss causation could be established merely by showing that the price of the stock on the date of purchase was inflated by the defendant's misrepresentations. The Supreme Court held that loss causation requires more, finding that Congress intended “to permit private securities fraud actions for recovery where, but only where, plaintiffs adequately allege and prove the traditional elements of causation and loss.” Id. at 346, 125 S.Ct. 1627. Thus, plaintiffs must “prove that the defendant's misrepresentation (or other fraudulent conduct) proximately caused the plaintiff's economic loss.” Id. Stated differently, the plaintiff must show a “causal connection” between the “loss and the misrepresentation.” Id. at 347, 125 S.Ct. 1627.

The parties and the Ninth Circuit agree on this much: that Plaintiffs must prove a causal connection between Defendants' fraudulent actions and their loss. The question is how that connection must be proved. On this question, the parties and the Ninth Circuit cases diverge.

B. Daou

and its Progeny.

Shortly after the Supreme Court decided Dura, the Ninth Circuit issued an amended opinion in Daou, 411 F.3d at 1006. The district court in Daou had dismissed the plaintiff's third amended complaint because it did not “allege that there were any negative public statements, announcements or disclosures at the time the stock price dropped that Defendants were engaging in improper accounting practices.” Id. at 1026. In other words, because the defendants' fraud—the improper accounting practices—had not been publicly disclosed, the district court concluded that loss causation had not been pled. The Ninth Circuit reversed. It observed that “the price of Daou's stock fell precipitously after defendants began to reveal figures showing the company's true financial condition.” Id. (emphasis added). The Ninth Circuit found loss causation to be adequately pled because “Plaintiffs allege that these disclosures of Daou's true financial health, the result of prematurely recognizing revenue before it was earned, led to a ‘dramatic, negative effect on the market, causing Daou's stock to decline[.]’ ” Id. (emphasis added). In other words, it was the disclosure of the company's financial problems—problems caused by the fraudulent accounting practices—that led to the stock decline and the plaintiff's loss.

That it was the disclosure of the company's financial condition, rather than disclosure of defendants' fraud, that satisfied loss causation, is made abundantly clear in Daou. The opinion on page 1026 refers to disclosure of “the company's true financial condition” and “Daou's true financial health.” Id. The next page refers to the disclosure of “Daou's true financial health,” “the true nature of Daou's financial condition,” and “Daou's true financial situation.” Id. at 1027. Although it is correct that the facts in Daou also included the revelation of additional information from which one market analyst became suspicious that the company was “manufacturing earnings,” and although it is also correct that the company disclosed a growing amount of unbilled receivables in one of its accounts, it was not the disclosure of these facts that the Ninth Circuit found sufficient for loss causation. Rather, it was the disclosure of the company's true financial condition, which had been previously misrepresented by the defendants, which led to a drop in the stock price and provided the causal connection between the defendants' wrongful conduct and the plaintiffs' loss. As the Ninth Circuit observed, “the price of Daou's stock fell precipitously after defendants began to reveal figures showing the company's true financial condition.” Id. at 1026.

The Ninth Circuit took the same approach three years later in Berson v. Applied Signal Technology, Inc., 527 F.3d 982 (9th Cir.2008). The plaintiffs in Berson bought stock in Applied Signal during the six months before the company revealed that its revenue had fallen 25%. Immediately following this disclosure, the stock price dropped 16% and plaintiffs sued the company and two of its officers for securities fraud. Id. at 984. The plaintiffs alleged that the company engaged in a misleading process of reflecting the dollar value of government contracts in a “backlog” account, suggesting that the company would perform the contracted-for work in the future and would earn the contracted-for revenues. Defendants did not disclose that some of those contracts were the subjects of “stop-work orders” from the government that meant they might never be performed. Thus, plaintiffs were given the incorrect impression that the company had a substantial backlog of future work, when in fact tens of millions of dollars in the backlog were under stop-work orders and might never be performed.

The Ninth Circuit provided this description in finding that the plaintiffs adequately pled loss causation: “The complaint describes the stop-work orders in detail, explains that the orders halted a significant amount of work, alleges that the reduced workload caused revenue to fall by 25%, and claims that this revenue reduction caused the stock price to drop by 16%.” Id. at 989 (emphasis added). In other words, it was the eventual effect of the misrepresented facts—the contracts subject to stop-work orders—that caused revenue to drop, stock prices to fall, and plaintiffs' injuries. The very facts that were wrongly withheld ultimately led to the plaintiffs' loss. As in Daou, it was the revelation of the company's true financial condition, in contrast to the misleading financial condition portrayed by the defendants, that led to the stock price drop and satisfied loss causation.

This approach became even clearer when the Ninth Circuit articulated this test for loss causation in Nuveen: “A plaintiff can satisfy loss causation by showing that ‘the defendant misrepresented or omitted the very facts that were a substantial factor in causing the plaintiff's economic loss.’ ” 730 F.3d at 1120 (emphasis in original) (citing McCabe v. Ernst & Young, LLP, 494 F.3d 418, 425 (3d Cir.2007)). Thus, drawing a causal connection between the facts misrepresented and the plaintiff's loss will satisfy loss causation. A plaintiff need not show that the fraudulent practices themselves were revealed: “Disclosure of the fraud is not a sine qua non of loss causation, which may be shown even where the alleged fraud is not necessarily revealed prior to the economic loss.” Id. at 1120.

The Nuveen test accurately describes the holdings in Daou and Berson. The “very facts” misrepresented in Daou—the company's earning capacity—ultimately led to lower revenues, the drop in stock price, and the plaintiff's loss. The “very facts” concealed in Berson—that several of the company's large contracts were subject to stop-work orders—ultimately led to the drop in revenue that produced the drop in stock price.⁷

7 As Defendants note, Nuveen is not a fraud-on-the-market case. The securities at issue in Nuveen were purchased in private transactions. Although Defendants argue that this fact distinguishes Nuveen from the present case, Nuveen itself explained that the loss causation test is the same for efficient and inefficient markets: “Although Nuveen repeatedly promotes a different standard for Rule 10b-5 claims arising from ‘inefficiently traded’ securities, the need to reliably distinguish among the tangle of factors affecting a security's price is no less urgent in efficient markets. ‘[F]undamentally, the same loss causation analysis occurs in both typical and non-typical § 10(b) cases.’ ” Id. at 1123 (quoting McCabe, 494 F.3d at 425 n. 2). The footnote from McCabe cited in Nuveen holds that the loss causation test is the same for stock purchased in publicly-traded (efficient) markets and stock purchased in private transactions. See 494 F.3d at 425 n. 2.

In summary, as the Court reads Daou, Berson, and Nuveen, proof of loss causation is not confined to a particular kind of market disclosure. The question is whether the facts misrepresented or concealed by the defendant led to the plaintiff's loss. If they did, then the defendant's misrepresentation or omission has a causal connection to the plaintiff's loss as required by Dura.

This rule is not, as Defendants contend, a form of investor insurance. The test does not establish a system under which a plaintiff, once having purchased stock, is protected against any and all possible losses. The only losses for which a plaintiff can recover are those caused by “the very facts” that were misrepresented or omitted. To use the Huddleston example quoted above, the investor in the ship could recover nothing if the loss was caused by the lack of insurance. But if the loss was due to the very facts that were misrepresented—the ship's carrying capacity—then the misrepresentation would be causally connected to the loss and proximate causation would be satisfied.

C. Metzler

and its Progeny.

Another line of Ninth Circuit cases takes a more restrictive view of loss causation. This line of cases appears to begin with Metzler Inv. GMBH v. Corinthian Colleges, Inc.. Purporting to apply Daou, Metzler concluded that to allege loss causation “the complaint must allege that the practices that the plaintiff contends are fraudulent were revealed to the market and caused the resulting losses.” 540 F.3d 1049, 1063 (9th Cir.2008) (emphasis added). A plaintiff must show “that the market learned of and reacted to [the] fraud, as opposed to merely reacting to reports of the defendant's poor financial health generally.” Id.

Respectfully, the Court regards this as a misreading of Daou. As noted earlier, Daou emphasized that the disclosure which triggered the plaintiff's loss and satisfied the requirement of loss causation was “the company's true financial condition.” 411 F.3d at 1026. Because that poor financial condition resulted from the very facts the defendants had misrepresented by prematurely recording revenues, loss causation was satisfied.

Despite this apparent misreading of Daou, the holding in Metzler has spawned additional cases. In Oracle, the Ninth Circuit made the holding in Metzler even clearer: “[L]oss causation is not adequately pled unless a plaintiff alleges that the market learned of and reacted to the practices the plaintiff contends are fraudulent, as opposed to merely reports of the defendant's poor financial health generally.” 627 F.3d at 392. In other words, the market must learn of the specific fraudulent practices. It is not enough that a plaintiff suffers loss because the very facts that were the subject of those fraudulent practices caused his loss. Even though Daou specifically stated—five times—that stock losses caused by revelation of the company's true financial condition can satisfy loss causation if that financial condition is caused by the misrepresented facts (411 F.3d at 1026–27), and even though Berson and Nuveen adopt the same approach, Oracle specifically states that plaintiffs cannot prove loss causation “by showing that the market reacted to the purported ‘impact’ of the alleged fraud—the earnings miss—rather than to the fraudulent acts themselves.” 627 F.3d at 392.

Oracle was followed by Loos v. Immersion Corp, 762 F.3d 880 (9th Cir.2014), and Oregon Public Employees Retirement Fund v. Apollo Group, Inc., 774 F.3d 598 (9th Cir.2014), both of which also held that loss causation requires proof that the company's fraudulent practices, as opposed to the adverse financial impact of those practices, was revealed to the market. In the Court's view, Metzler, Oracle, Loos, and Apollo adopt a more restrictive view of loss causation than Daou, Berson, and Nuveen. Securities fraud plaintiffs can recover only if the market learns of the defendants' fraudulent practices. It is not enough that plaintiffs are injured by the consequences of those practices.⁸

8 Another securities fraud case decided by the Ninth Circuit during this same general time period, In re Gilead Sciences Securities Litigation, 536 F.3d 1049 (9th Cir.2008), does not clearly embrace any single approach to proving loss causation. Gilead was a case where the defendant's improper off-label marketing was revealed to the market and, later, when combined with a revenue drop, resulted in loss to the plaintiffs. It illustrates that loss causation can in fact be proved in the way Metzler and its progeny require, but does not suggest that is the only way loss causation can be established. The Court notes that Gilead cites favorably to the Third Circuit's decision in McCabe from which the Nuveen loss causation test is drawn. See id. at 1057.

The Court pauses to address a concern that may underlie Metzler and these later cases—that recognizing loss causation merely from a company's poor financial health may lead to the recovery of losses that were caused by factors other than the defendant's fraud. The Court agrees that such a rule would be an improper form of investor insurance, but that is not what Daou, Berson, and Nuveen permit. They require the plaintiff to prove more than the company's poor financial health and a resulting stock drop. The plaintiff must also prove that the company's poor financial health was caused by the “very facts” that the defendant misrepresented or concealed. The plaintiff clearly must prove a causal connection between the fraud and the loss.

D. Which Line of Cases Should the Court Follow?

The Court concludes that it should apply the loss causation test adopted in Daou, Berson, and Nuveen. It reaches this conclusion for three reasons.

First, Daou was decided before any of the other cases. Because all of the cases discussed above were decided by three-judge panels of the Ninth Circuit, none of those panels had authority to overrule Daou. See Galbraith v. Cnty. of Santa Clara, 307 F.3d 1119, 1123 (9th Cir.2002) (“[A] three judge panel normally cannot overrule a decision of a prior panel on a controlling question of law[.]”). Applying this principle, courts generally hold that when two panel decisions conflict, the earlier panel decision controls. See McMellon v. United States, 387 F.3d 329, 333 (4th Cir.2004); Wilson v. Coman, 284 F.Supp.2d 1319, 1339 (M.D.Ala.2003). Because Daou is the earlier panel decision, the Court will follow it.

Second, the Court views the Daou line of cases as stating the better rule. As explained in Dura and explored more thoroughly in McCabe, loss causation is a form of proximate cause. It was adopted by Congress to ensure that securities fraud plaintiffs may recover from defendants only when the actions of those defendants proximately cause the plaintiffs' losses. Dura, 544 U.S. at 342–46, 125 S.Ct. 1627. Such causation is assuredly established when the “very facts” misrepresented or concealed by the defendant cause the plaintiff's loss. For example, if a company publicly overstates its manufacturing capacity, a plaintiff purchases stock at an inflated value because of the company's misrepresentation, and the plaintiff's stock later loses value because a competitor's newly-developed product eclipses the company's product and causes a drop in the company's revenues, loss causation has not been satisfied. The development of a better competing product, not the fact misrepresented by the company (manufacturing capacity), led to the plaintiff's loss. If, however, the company's revenues fail to meet projections because of the lack of manufacturing capacity—the very fact misrepresented—and the stock loses value as a result, the misrepresented fact has led to the plaintiff's loss. This is true even if the market does not learn that the company lied. If the plaintiff can prove that the drop in revenue was caused by the misrepresented fact and that the drop in his or her stock value was due to the disappointing revenues, the plaintiff should be able to recover. A causal connection between the “very fact” misrepresented and the plaintiff's loss has been established.

Third, traditional notions of proximate cause are not so narrowly circumscribed as the rule in Metzler and its progeny. Section 548A of the Restatement (Second) of Torts, which the Supreme Court in Dura described as “setting forth the judicial consensus,” 544 U.S. at 344, 125 S.Ct. 1627, provides this relevant explanation of loss causation (referred to in the Restatement as “legal causation”):

Thus one who misrepresents the financial condition of a corporation in order to sell its stock will become liable to a purchaser who relies upon the misinformation for the loss that he sustains when the facts as to the finances of the corporation become generally known and as a result the value of the shares is depreciated on the market, because that is the obviously foreseeable result of the facts misrepresented. On the other hand, there is no liability when the value of the stock goes down after the sale, not in any way because of the misrepresented financial condition, but as a result of some subsequent event that has no connection with or relation to its financial condition.

Restatement (Second) of Torts § 548A, Comment b (1977) (emphasis added). This traditional rule does not require that the fraud become known, only that the “facts as to the finances of the corporation” become known. This precisely describes the holdings in Daou and Berson.

For these reasons, the Court will follow Daou, Berson, and Nuveen. The loss causation test the Court will apply is this: “A plaintiff can satisfy loss causation by showing that the defendant misrepresented or omitted the very facts that were a substantial factor in causing the plaintiff's economic loss.” Nuveen, 730 F.3d at 1120 (emphasis in original; citation omitted). The fraud or misrepresentation “need not be the sole reason for the decline in value of the securities, but it must be a substantial cause.” Gilead, 536 F.3d at 1055 (internal quotation marks omitted).

E. Section 1292(b) Certification.

For the reasons set forth below, the Court finds that application of the Daou loss causation test largely results in denial of Defendants' motion for summary judgment. Had the Court applied Metzler and its progeny, Defendants' motion would be granted in full because Plaintiffs have not presented evidence from which a reasonable jury could find that Defendants' alleged fraudulent practices became known to the market during the class period.

Denial of Defendants' motion will result in the parties embarking on expensive expert discovery and a costly and complex trial, none of which will be necessary if the Ninth Circuit concludes that Metzler and its progeny represent the correct loss causation test. To avoid this potentially unnecessary expense for the parties and the Court, the Court will take the unusual step of certifying the loss causation issue for immediate interlocutory appeal. The Court concludes that the loss causation test is a “controlling question of law as to which there is substantial ground for difference of opinion and that an immediate appeal from [this] order may materially advance the ultimate termination of the litigation.” 28 U.S.C. § 1292(b). The issue certified is this: what is the correct test for loss causation in the Ninth Circuit? Can a plaintiff prove loss causation by showing that the very facts misrepresented or omitted by the defendant were a substantial factor in causing the plaintiff's economic loss, even if the fraud itself was not revealed to the market (Nuveen, 730 F.3d at 1120), or must the market actually learn that the defendant engaged in fraud and react to the fraud itself (Oracle, 627 F.3d at 392)? Within 10 days of the entry of this order, either side may petition the Ninth Circuit to decide this issue on immediate appeal. Id. If neither side files such a petition, the Court will schedule a case management conference to set a schedule for completion of expert discovery and clarification of issues (as discussed below), and will set a final pretrial conference, at which a firm trial date will be set. If a petition for immediate appeal is filed within 10 days, the Court will stay this action until the Ninth Circuit decides whether to take the appeal and, if it does, the stay will remain in effect until the appeal is decided.

IV. Loss Causation Analysis.

In this section, the Court will address whether Plaintiffs have presented evidence from which a reasonable jury could find the Daou loss causation test satisfied. The Court will address other § 10(b) issues in the next section.

One of the difficulties presented by this case arises from the parties' failure to agree on precisely which alleged misrepresentations and omissions form the basis for Plaintiffs' case. Defendants identify 167 false statements from Plaintiffs' complaint. Doc. 327-1 at 1-254. Plaintiffs identify 96 false statements. Doc. 363 at 80-121. In addition to being different in number, the identified statements differ in content. Defendants ask the Court to enter summary judgment on a misrepresentation-by-misrepresentation basis, but Plaintiffs do not address Defendants' list, much less attempt to identify the evidence that underlies each misrepresentation on the list. Nor do Plaintiffs attempt to present specific evidence for the false statements on their list. Plaintiffs instead take a more general approach, arguing that six events “removed the price inflation caused by defendants' earlier misstatements and omissions” about the LPM and hot climate defects. Doc. 363 at 65. Plaintiffs claim that each event caused the price of First Solar stock to decline due to the revelation of increased costs that had previously been concealed by Defendants. Plaintiffs rely heavily on the declaration of their expert, Bjorn I. Steinholt, who analyzes whether costs related to the LPM and hot climate defects substantially contributed to First Solar's poor financial health and resulting stock declines. Doc. 374.

Defendants argue that the Court should grant summary judgment on each of their 167 misrepresentations because Plaintiffs have not addressed them individually and identified the evidence from which a reasonable jury could find that they were made and caused Plaintiffs' losses. But the Court cannot conclude that Plaintiffs must prove their case as Defendants configure it, or even that Defendants' list of misrepresentations accurately reflect Plaintiffs' case. Neither can the Court determine that Plaintiffs' list of 96 misrepresentations is correct—as noted, Plaintiffs make no attempt to identify the evidence that supports them or show that they caused Plaintiffs' loss. The parties clearly failed to communicate about the issues to be addressed in the motion and response, but the Court cannot conclude that this lack of clarity provides a basis for summary judgment. As discussed below, Plaintiffs have identified evidence supporting their claim that the facts allegedly misrepresented and omitted by Defendants affected the company's financial health and caused Plaintiffs' losses. The Court finds that this evidence, if accepted by a jury, could satisfy the Daou loss causation test.

The Court remains concerned, however, about the lack of clarity in this case approaching trial. The Court could hold that Plaintiffs are limited to proving the six events addressed in their brief, but the evidence they cite and rely on goes well beyond those events. The Court could require the parties to agree on a list of misrepresentations and omissions and redo the summary judgment briefing, but this would require substantial additional time and expense for the parties and the Court. The Court could rely on the final pretrial report to identify the precise issues to be addressed at trial, but this too will almost certainly spawn disagreements. The Court feels considerable frustration over this state of affairs and concludes that the case must be clarified before trial, but also concludes that this is a matter to be addressed after the § 1292(b) appeal is resolved. As noted, that appeal could result in the Court granting summary judgment for Defendants even in light of the evidence presented by Plaintiffs. If that is not the result, the Court and the parties will have to figure out how to prepare this currently-confused case for trial.⁹

9 Another problem arises from Plaintiffs' heavy reliance on their experts in the summary judgment briefing. The parties agreed that expert discovery would occur after summary judgment briefing and would not provide a basis for further summary judgment motions. See Doc. 177, ¶ 5 (“Expert discovery shall occur after the Court's final ruling on motions for summary judgment, and shall not provide a basis for additional summary judgment motions.”). Because expert discovery has not occurred, Defendants argue that Plaintiffs should be precluded from relying on their experts in opposing summary judgment. But the Court clearly cannot grant summary judgment in disregard of expert opinions that would be available at trial. The Court therefore will consider Plaintiffs' expert submissions.

A. July 29, 2010—Earnings Release.

On July 29, 2010, First Solar announced its 2Q10 earnings and disclosed the LPM defect for the first time. Doc. 374, ¶ 35. First Solar beat Bloomberg consensus estimates on its earnings per share and revenues. Id., ¶ 34. It also reduced its revenue guidance by $100 million and increased its earnings per share guidance for fiscal year 2010 from $6.80-$7.30 per share to $7.00-$7.40 per share. Id. First Solar estimated that the LPM would negatively impact its revenues by $99 million. Id.

That day, First Solar also held a conference call to discuss the LPM defect, during which Gillette made the following statement:

Finally in Q2, reflected costs associated with the modular replacement program. During the period from June of 2008 to June of 2009, a manufacturing excursion occurred affecting less than 4% of the total product manufactured within the period. The excursion could result in possible power loss in affected modules. The root cause was identified and subsequently mitigated in June of 2009. Ongoing testing confirms the corrective actions are effective. We have been working directly with impacted customers to replace the affected modules and these efforts are well under way, and in some cases complete. We accrued the estimated full cost of these additional efforts in our Q2 results and Jens will discuss the financial impact in more detail.

Id., ¶ 35. Jens Meyerhoff also commented on the LPM defect:

During the second quarter, we accrued $17.8 million in cost of sales for expected module replacement costs and our cost of goods sold. In addition, we accrued $5.6 million of operating expenses associated with this process excursion, bringing our total accrued expenses to 27.4 million at the end of the second quarter.

Id., ¶ 36.

Market analysts had positive comments about beating consensus estimates, but also noted the lower revenue guidance. Cantor Fitzgerald stated: “The company is capacity constrained....This results in a drop in total revenue guidance, but a slight increase in earnings guidance. We expect that most investors will find this disappointing.” Id., ¶ 39. Needham noted: “The company raised its full year earnings guidance, but lowered its 2010 revenue outlook, which probably disappointed the Street given the high expectations going into the report, in our view.” Id.

Analysts also commented on the LPM problem. Credit Suisse reported the new accruals and stated, “bears will point out why the charge is being taken more than a year after the company knew about and resolved the issue, and question why a similar issue will not arise again.” Id., ¶ 40. UBS noted: “Potentially concerning takeaways from its 2Q10 results....We believe this could be an overhang on the stock as the modules are replaced over the next six months.” Id.

Following the July 29 disclosure, First Solar stock decreased by $10.05 per share, or 7.4%. Id., ¶ 42. In an internal email, First Solar noted investor concerns and recognized that the LPM defect caused the stock drop: “Excluding [the LPM defect] we would have achieved [the expected financial numbers]. [The LPM problem] tarnished our flawless execution image. Rumors that we will incur more costs than the Q2 charge. Some fear why not more than 4% and customers saying bigger problem.” Id., ¶ 41.

Steinholt also concludes that the LPM problems caused the stock drop: “Absent the LPM problems, First Solar would have reported an estimated 13% higher earnings for 2Q10, and not have had to reduce its 2010 revenue guidance by $100 million.” Id. The “LPM expenses and approximately $100 million lost revenues related to the LPM problem explain all, or at least a substantial portion, of the Company-specific stock price decline on July 30, 2010.” Id. In other words, Steinholt opines that the “very facts” allegedly omitted by Defendants—the existence of the LPM defect—ultimately led to a drop in stock price that caused Plaintiffs' loss. A reasonable jury could find from this evidence that Plaintiffs' have proved loss causation under Daou.

B. February 24, 2011—Earnings Release.

On February 24, 2011, First Solar announced 4Q10 earnings results, beating the Bloomberg consensus estimate for earnings per share but missing on estimated revenues. Id., ¶ 43. First Solar decreased the high end of its revenue guidance from $3.7-$3.9 billion to $3.7-$3.8 billion and increased its earnings per share guidance from $8.75-$9.50 per share to $9.25-$9.75 per share. Id. It noted additional accrued expenses of $8.5 million for the LPM defect. In the related conference call, Gillette explained that “Q4 was impacted by our decision to divert some volumes to expedite the module replacement program,” which was also confirmed by Zhu. Id., ¶ 44.

The next day, analysts commented on the figures, most notably the revenue miss. Mizuho Securities noted: “[R]evs missed guidance somewhat ... due largely to a decision to accelerate replacement of ~30MW of potentially faulty modules.” Id., ¶ 45. Auriga stated: “We expect bears to raise the issue of revenue falling short of the consensus in both 4Q10 and 1Q11.” Id. Ardour stated: “4Q10 revenues somewhat light, but EPS continues to outperform.” Id.

After the disclosure, First Solar stock dropped $8.96 per share, or 5.4%. Id., ¶ 47. But for the LPM problems, Steinholt opines, “First Solar would have beat Bloomberg 4Q2010 EPS consensus by 16 cents (as opposed to 7 cents) and avoided reporting a $37 million 4Q2010 revenue miss.” Id. Steinholt concludes that “LPM expenses and the lower than expected revenues explains all, or a substantial portion, of the [stock decline].” Id. Given this evidence, a reasonable jury could find that the very facts Defendants allegedly fraudulently concealed—the scope of the LPM defect and its resulting financial impact— were substantial factors in causing Plaintiffs' loss.

C. May 3, 2011—Earnings Release.

On May 3, 2011, First Solar issued its numbers for 1Q11, beating Bloomberg consensus estimates for earnings per share and revenues. Id., ¶ 48. First Solar also announced additional expenses of $4.5 million for the LPM defect and maintained its revenue and earnings per share guidance for fiscal year 2011. Id., ¶¶ 48-49. Operating income was reduced by $10 million and guidance for operating cash flow was reduced. Id., ¶ 49.

First Solar's guidance had accounted for some impact due to the heat degradation issue, but this fact was not disclosed in the financial statements. Id. “Unbeknownst to investors, First Solar's 2011 guidance now incorporated some impact from the heat degradation problem.” Id. Steinholt further notes that “[b]ecause the heat degradation issue was not specifically discussed, or otherwise disclosed or broken out, none of the analysts had an opportunity to comment on the issue in subsequent analyst reports.” Id., ¶ 50. Analysts did express “disappointment that First Solar did not raise guidance.” Id.

First Solar stock dropped by $8.35 per share, or 6.2%. Id., ¶ 51. Steinholt opines that “1Q2011 LPM expenses and the impact of the heat degradation issue on the Company's 2011 guidance had a negative impact on First Solar's stock price, and, therefore, contributed to its May 4, 2011 stock price decline.” Id. He also notes that the “heat degradation problem negatively impacted reported 2011 revenue guidance by $24 million,” resulting in a “$0.20 per share hit to guidance.” Id., ¶ 49. Steinholt concludes that this “would imply a $2.66 per share negative impact” to First Solar's stock price. Id.

Unlike the two prior releases, Steinholt does not suggest that the LPM defect and hot climate degradation “explain[ed] all, or a substantial portion” of the stock decline. Nevertheless, Plaintiffs need not show “ ‘that a misrepresentation was the sole reason for the investment's decline in value’ in order to establish loss causation. ‘[A]s long as the misrepresentation is one substantial cause[,] other contributing forces will not bar recovery’ but will play a role ‘in determining recoverable damages.’ ” Daou, 411 F.3d at 1025 (quoting Robbins v. Koger Props., Inc., 116 F.3d 1441, 1447 n. 5 (11th Cir.1997)). A reasonable jury could determine that the very facts omitted and misrepresented by Defendants—the effect of the LPM defect and existence of the hot climate degradation issue—were substantial factors in causing the stock to decline and Plaintiffs' loss.

D. October 25, 2011—Gillette Leaves First Solar.

On October 25, 2011, Gillette was replaced as CEO. Doc. 374, ¶ 52. First Solar issued a press release that stated the following:

The Board of Directors of First Solar, Inc. (NASDAQ: FSLR) today asked its Chairman and company founder, Michael Ahearn, to serve as interim Chief Executive Officer. Ahearn has accepted. Effective immediately, Rob Gillette is no longer serving as Chief Executive Officer, and the Board of Directors thanks him for his service to the company. The Board of Directors has formed a search committee and is initiating a search for a permanent Chief Executive Officer.

Id. Several analysts commented on the departure of Gillette. Credit Suisse stated:

The news is clearly negative...the abruptness of the announcement and the terse wording of the release, the fact that earnings will likely be next week and the CEO is stepping down just a week in advance, and the fact that FSLR had planned to host a reception with the CEO on Nov 15 all sound unfortunately quite concerning.

Id., ¶ 53. Goldman Sachs stated: “[T]he terse nature of today's announcement, its timing and the lack of a permanent replacement argue that this was an unanticipated event.” Id. Deutsche Bank stated: “The press release is short on details and we believe the news (along with the timing of the announcement) is likely to raise a lot of investor questions about the health of overall industry as well as near/longer term profitability outlook of the company.” Id. Raymond James stated: “So, what could have prompted this sudden change?...On the bearish side would be an accounting scandal....A less damaging but still negative scenario would be the board sacking Gillette ahead of a major earnings miss or guidance cut.” Id. Canton Fitzgerald, Morgan Stanley, PacificCrest, Wunderlich, and Jeffries echoed similar sentiments. Id.

Immediately following this news, First Solar's stock price declined $14.48 per share, or approximately 25%. Id., ¶ 54. Steinholt opines that “First Solar's press release disclosing [Gillette's] departure clearly was the reason” for the stock decline. Id. The next day, First Solar issued its 3Q11 results, but did not hold a conference call. Id., ¶ 55. The stock price rebounded by $2.84 per share on October 25 and $6.79 per share on October 26, 2011. Id., ¶ 57.

The October 25 press release that caused the stock price to drop did not include any information about the company's financial performance, the LPM defect, or the hot climate degradation issue. No financial statements were released that revealed additional financial impacts from the LPM or hot climate defects. Thus, Plaintiffs have failed to present evidence that the stock price decline was caused by Defendants' misrepresenting or omitting information about those two defects. The press release concerned Gillette's departure, and Plaintiffs have presented no evidence that he left First Solar because of Defendants' alleged fraud. Although some market analysts speculated that Gillette's termination could be due to internal company problems, such speculation was itself not related to the facts allegedly misrepresented and omitted by Defendants. Plaintiffs therefore have failed to present evidence from which a reasonable jury could find that their loss was caused by the facts allegedly misrepresented or omitted by Defendants.

The Court rejects Plaintiffs' argument that a jury could simply infer a connection between Gillette's departure and the alleged fraud. Plaintiffs have provided no evidence that Gillette left First Solar because of the alleged fraudulent activity, and any such inference would be based on pure speculation. The Court will enter summary judgment with respect to the stock decline on October 25, 2011. See Celotex, 477 U.S. at 322, 106 S.Ct. 2548.¹⁰

10 This conclusion provides an illustration of how the test in Daou, Berson, and Nuveen does not conflate transaction and loss causation. Plaintiffs may be able to show that Defendants' alleged misrepresentations caused them to purchase First Solar stock at a particular price, but they cannot show that those misrepresentations caused the losses resulting from Gillette's departure. As a result, they cannot recover for those losses.

E. December 14, 2011—Guidance Updates.

On December 14, 2011, First Solar reduced its earnings guidance from $6.50-$7.50 per share to $5.75-$6.00 per share and reduced revenue guidance from $3.0-$3.3 billion to $2.8-$2.9 billion, missing Bloomberg consensus estimates. Doc. 374, ¶ 59. First Solar also announced restructuring charges of $0.85 per share during 4Q11, which included eliminating 100 positions. Id. First Solar updated its 2012 guidance for earnings per share and revenues, both of which fell below Bloomberg's consensus estimates. Id., ¶ 60. In addition, Ahearn stated in the press release that First Solar was “recalibrating our business to focus on building and serving sustainable markets rather than pursuing subsidized markets.” Id.

First Solar stock fell $9.12 per share, or 21.4%. Id., ¶ 61. Steinholt opines that First Solar's “reduced 2011 guidance, and initial 2012 guidance significantly below expectations, explains the Company-specific portion of First Solar's December 14, 2011, price decline.” Id. Steinholt attributes the low guidance to the “heat degradation problem as the Company was changing its focus to larger scale projects in hotter climates.” Id., ¶ 60. From this evidence, a reasonable jury could conclude that the facts omitted by Defendants relating to the hot climate defect revealed the true financial condition of First Solar and were a substantial factor in the stock price decline.

F. February 28, 2012—Earnings Release.

On February 28, 2012, First Solar announced its 4Q11 numbers. Id., ¶ 62. These reflected substantial losses, which included “(a) a non-cash goodwill impairment charge of $3.90 per share; (b) restructuring charges of $0.43 per share (below $0.85 per share announced on December 14, 2011); and (c) $1.67 related to warranty and cost in excess of normal warranty expense, for a total of $6.00 per share.” Id. First Solar met the low end of its revenue guidance for fiscal year 2011. Id. Revenue and cash flow guidance for 2012 were also lowered. Id., ¶ 63.

In the conference call, Ahearn commented about the additional warranty expenses:

This quarter we incurred $125.8 million in additional warranty reserves to reflect an updated estimate of costs related to the manufacturing excursion that occurred between June 2008 and June 2009. As previously disclosed, a small percentage of product manufactured during that time period may experience premature power loss once in the field. First Solar identified and addressed the manufacturing excursion in June 2009 and later initiated a voluntary remediation program that goes above and beyond our standard warranty obligations. The remediation program includes module removal, testing, replacement and logistical services and additional compensation payments to customers under certain circumstances.

A large volume of claims made under the remediation program were processed in the fourth quarter, and we identified the significant increase in remediation costs under the terms of our voluntary program. Our estimates now benefit from having processed over 95% of the total claims submitted under the life of the program. The total cost of remediating the manufacturing excursion that occurred from June 2008 to June 2009 now stands at $215.7 million including $145.6 million above and beyond our standard warranty.

There are approximately 4% of the claims submitted for which we have not yet been able to determine if remediation is required. If it is determined that these claims should be remediated, there's at least the potential for additional costs of as much as $44 million.

Id., ¶ 64. Widmar broke down the additional costs during the call:

The first item is the cost to remove, replace and provide logistical services related to the manufacturing excursion. In the fourth quarter we expensed $23.9 million for these efforts and have expensed $99.7 million to date.

The second item is expected payment to customers under certain conditions for power loss prior to the remediation of the customer's system. In the fourth quarter we expensed $31.8 million for this compensation and have expensed $45.9 million to date.

The third item, $70.1 million, is due to an increase in the expected number of replacement modules above our standard warranty rate required for our remediation efforts.

Id., ¶ 65. He then discussed the hot climate degradation:

Finally, we recognized a $37.8 million charge to increase our warranty accrual. We believe our PV modules are potentially subject to increased failure rates in hot climates. As our geographic mix of sales has shifted to hot climates we have increased our warranty accrual. Our experience has shown that our warranty rate for hot climates are slightly higher than the return rates for temperate climates. With this change, our standard warranty accrual rate has been increased by one percentage point to account for potential returns going forward. We will continue to review our warranty accrual rate in the future and will adjust the rate as appropriate to reflect our actual experience.

Id.

During the question and answer portion of the call, an analyst noted that Ahearn had previously said that all the warranty expenses had been taken into account in prior financials. He then asked what had changed from November 3 to the end of the quarter. Id., ¶ 66. Ahearn responded:

[W]e processed a large volume of the claims that were made over the life of the program in Q4. For the last quarter we reserved and reported based on the best available information then and we discovered in processing the claims a lot of additional exposure, which is reflected in the charges that we've taken in Q4.

Id. Another analyst asked about the hot climate degradation issue and requested Ahearn to give some quantitative data about the problem. Id., ¶ 67. Ahearn responded that First Solar lacked data but that it was taking a conservative approach to the warranty rate and would continue to reevaluate as it gathered more results. Id.

Analyst reaction to the earnings misses was tempered. Ardour noted: “We believe the 4Q11 miss was somewhat expected.” Id., ¶ 69. Deutsche Bank, Goldman Sachs, and Morningstar noted that 2012 guidance was “intact.” Id. But analysts expressed more concern over the LPM defect and hot climate degradation issues, noting that investors should be wary of additional future warranty accruals, especially considering First Solar had previously stated that it believed that bulk of warranty risk was behind it. Id., ¶ 70.

On February 29, 2012, First Solar filed its 2011 Form 10-K disclosing that “the Company had taken a $13.8 million module inventory write-down primarily as a result of the voluntary remediation efforts.” Id., ¶ 71. Credit Suisse released an analyst report addressing “credibility and brand concerns that now likely exist with investors and customer partners” as a result of the ever-increasing costs attributed to the LPM defect. Id. The report also stated that additional charges could come in the future, and noted the hot climate degradation issue. Id.

Gordon Johnson of Axiom commented about First Solar's situation on CNBC:

Their stuff is not working in the field. That's effectively what's happening. Forget about the fact that they massively missed earnings. This is a huge red flag. It brings into question whether they will be able to do projects here in the U.S. This is a game changer.

We heard from our checks in Germany that this is not a one-time issue. And the fact that banks are becoming cautious on lending to First Solar projects suggests that there is a fundamental problem with their modules. ...This is new. This is huge, and it is potentially going to be a game ender.

This has been a problem that First Solar has had in the past, and as was asked on the call by one of our competitors, they told us that this problem was, you know, was fixed. And it is not fixed. So, we need to look into this further. The company will not talk to us. So, we definitely need to do some more checks to see that we are accurate. But at first glance, this is quite negative.

Id., ¶ 73.

First Solar's stock fell by $4.10 per share, or 11.26%, on February 29. Id., ¶ 74. The next day, it fell an additional 5.8%. Id. Steinholt states “that the disclosures relating to the LPM and heat degradation issues explains all, or at least a substantial portion, of the Company-specific stock price declines on February 29, 2012 and March 1, 2012.” Id. He notes that many costs related to the two defects were quantified and disclosed in the earnings release, including “(a) $1.67 per share for warranty and cost in excess of normal warranty expense, (b) additional $44 million in costs related to outstanding claims, or approximately $0.45 per share, and (c) $13.8 million in module inventory writedown, or $0.16 per share.” Id., ¶ 72. Steinholt notes that the impact of the two defects on future sales was “difficult to quantify precisely.” Id., ¶ 73. From this evidence, a reasonable jury could find that Defendants' alleged concealment of the true scope of the defects along with alleged accounting violations caused a negative financial impact to First Solar's sales, a drop in revenue and guidance, and Plaintiffs' losses.

G. Conclusion.

Plaintiffs have presented sufficient evidence to avoid summary judgment on loss causation with respect to five of the six alleged stock price declines. Plaintiffs have failed to do so for the stock decline that followed Gillette's departure on October 25, 2011. Defendants' motion will be granted with respect to the losses that occurred on that date, and denied with respect to the remaining drops.

V. Remaining § 10(b) Elements.

Defendants also challenge Plaintiffs' ability to prove that Defendants made material misrepresentations or omissions with scienter. These elements are closely intertwined and dependent upon similar facts.

In order to “fulfill the materiality requirement ‘there must be a substantial likelihood that the disclosure of the omitted fact would have been viewed by the reasonable investor as having significantly altered the ‘total mix’ of information made available.' ” Basic Inc. v. Levinson, 485 U.S. 224, 231-32, 108 S.Ct. 978, 99 L.Ed.2d 194 (1988) (quoting TSC Indus., Inc. v. Northway, Inc., 426 U.S. 438, 449, 96 S.Ct. 2126, 48 L.Ed.2d 757 (1976)). “Section 10(b) and Rule 10b-5(b) do not create an affirmative duty to disclose any and all material information.” Matrixx Initiatives, Inc. v. Siracusano, 563 U.S. 27, 131 S.Ct. 1309, 1321, 179 L.Ed.2d 398 (2011). “Silence, absent a duty to disclose, is not misleading under Rule 10-b-5.” Basic, 485 U.S. at 239 n.17, 108 S.Ct. 978. “Determining materiality in securities fraud cases ‘should ordinarily be left to the trier of fact.’ ” SEC v. Phan, 500 F.3d 895, 908 (9th Cir.2007) (quoting In re Apple Computer Secs. Litig., 886 F.2d 1109, 1113 (9th Cir.1989)).

Plaintiffs must also “prove that the defendant acted with scienter, ‘a mental state embracing intent to deceive, manipulate, or defraud.’ ” Tellabs, Inc. v. Makor Issues & Rights, Ltd., 551 U.S. 308, 319, 127 S.Ct. 2499, 168 L.Ed.2d 179 (2007) (quoting Ernst & Ernst v. Hochfelder, 425 U.S. 185, 193-94, 96 S.Ct. 1375, 47 L.Ed.2d 668 (1976)). A securities plaintiff may also establish scienter under § 10(b) by showing defendants acted with “deliberate recklessness.” In re Silicon Graphics Sec. Litig., 183 F.3d 970, 977 (9th Cir.1999). A “strong inference of ... ‘deliberate recklessness' ” is required. Id. “[T]he danger of misleading buyers must be actually known or so obvious that any reasonable man would be legally bound as knowing.” Hollinger v. Titan Capital Corp., 914 F.2d 1564, 1568–69 (9th Cir.1990).

Plaintiffs' allege several acts of fraud spanning nearly four years. They focus on the two general product defects mentioned above: the LPM problem and the hot climate degradation. Plaintiffs assert Defendants committed three acts of fraud with respect to each defect: (a) failing to disclose the defect to the public, (b) concealing the true scope of the problem from the public, and (c) intentionally underestimating the financial impact of the defect on financial statements. Plaintiffs also assert that Defendants manipulated an internal accounting metric, CpW, which was included in their SEC filings.

A. The LPM Defect.

1. Failure to Disclose.

In April 2008, Koralewski sent an email to Eaglesham regarding STBi data at First Solar's Perrysburg, Ohio plant: “FYI, not a real good trend and yes there are more than 10 data points in the monthly groupings.” Doc. 364-6 at 48. In 2Q09, Koralewski determined that approximately 2.9 million modules could potentially experience –20% STBi, which, according to Plaintiffs, demonstrated the severity of the LPM defect. Doc. 364-7 at 73. In June 2009, Eaglesham emailed Koralewski and stated that “[u]ntil we have a tighter window we should consider the possibility that 10% of the product produced in the last 10 months is affected.” Doc. 364-1 at 2. On May 31, 2009, Koralewski sent an email to another First Solar employee noting that “we have a rather serious quality problem that reared up late last week that has escalated over the weekend. ... We have a significant number of customers who are complaining about lower power modules and it looks real.” Doc. 364-7 at 116. Zhu and Sohn were also made aware of the defect. Docs. 364-8 at 2; 365-1 at 38. Ahearn began attending meetings addressing the status of the LPM defect. Doc. 364-7 at 100.

In July 2009, Ahearn was informed that “[t]he numbers are low, less than 5 percent of the total array, consisting of modules that are from the 159K population (potential LPM)....If this holds, the team will recommend no customer engagement as this will not be detectable.” Doc. 364-8. Ahearn responded that “[w]e'll have to keep our fingers crossed.” Id. Low power team meeting notes sent between Ahearn, Sohn, Meyerhoff, and Eaglesham stated that they “[m]ust be cautious not to leak unnecessary information either internally or externally.” Id. at 57. In preparation for the 2Q09 conference call, Sohn noted that “[t]here should not even be a question in the list about [LPM]. As far as the public is concerned, it does not exist....” Doc. 364-9 at 19. In October 2009, Meyerhoff wrote to Gillette that LPM is one of “the biggest smoking gun[s] we have at the company.” Doc. 364-1 at 7. In May 2010, Sohn told two First Solar employees that “[w]e should NOT be mentioning an ‘excursion’ publicly. E-staff has consciously made this decision....” Id. at 133 (emphasis in original).

Defendants argue that they had no duty to disclose the LPM defect because: (1) First Solar's alleged failure to disclose did not render other disclosed information misleading and (2) the defect did not become material until July 2010.

The first argument overlooks statements made prior to disclosure of the defect. For example, in a February 2009 conference call, an analyst asked Sohn to comment on “problems with the line this quarter...to make sure there was no manufacturing problems around, or that accounts for that line calc being down.” Doc. 372-1 at 402-03. Sohn appears to have dodged the question, responding instead that the holidays may have attributed to “line calc” being down and that First Solar was “very cautious and careful and [has] a high degree of expectation in terms of the way we operate the factories.” Id. at 403. In June 2009, Eaglesham told investors that “our track [record] of field performance and our knowledge of field performance allows us to have faith, high confidence that our product is delivering in the field.” Doc. 372-3 at 50. These statements were made when First Solar knew of the LPM problem and at least some internal estimates had suggested it was severe. See Dura, 544 U.S. at 341, 125 S.Ct. 1627.

Defendants' second argument—that the LPM defect did not become material until July 2010—is not one the Court can accept as a matter of law. In Matrixx, the Supreme Court considered whether “a company's failure to disclose reports of adverse events associated with a product” was material for purposes of securities fraud “if the reports do not disclose a statistically significant number of adverse events.” 131 S.Ct. at 1313. The plaintiffs brought suit after it was discovered that the company had received reports that one of its products, Zicam, may cause anosmia. Id. The Court found that even a non-statistically-significant number of adverse events could satisfy the materiality requirement because “medical professionals and regulators act on the basis of evidence of causation that is not statistically significant, [and] it stands to reason that in certain cases reasonable investors would as well.” Id. at 1321. Although companies do not have a duty to disclose all material information to investors, the adverse events reports in Matrixx were not merely anecdotal, but instead “plausibly indicated a reliable causal link between Zicam and anosmia.” Id. at 1322. The problem was material because it was likely that a reasonable investor would consider this information to have significantly altered the total mix of information made available. Id. at 1323.

Like the defendants in Matrixx, Defendants here argue that Koralewski's initial estimate of 75,000 modules in 2Q09, which resulted in a $1.8 million LPM Remediation Accrual, was less than 1% of First Solar's net income and thus the Disclosure Committee correctly deemed it non-material. They also argue Koralewski's 3Q09, 4Q09, and 1Q10 estimates, which represented 1.4%, 2.2%, and 2.6% of quarterly net income, respectively, were immaterial as well. In essence, they ask the Court to identify, as a matter of law, the percentage of quarterly net income at which the financial impact of a product defect becomes material. But the test for materiality is factual—whether a misrepresented or omitted fact would have been viewed by a reasonable investor as having significantly altered the total mix of information made available—and Defendants fail to provide undisputed evidence that a reasonable investor would consider a product defect irrelevant if it constitutes, say, only 1.5% of net income.

Moreover, Plaintiffs have presented evidence that First Solar knew the LPM defect was a serious problem that would concern investors well before it was disclosed. First Solar executives were careful not to release information about the LPM defect to the public, even deciding to delay informing customers until the customers discovered the problem themselves. Their evident concern about the potential market reaction to the LPM defect could be viewed by a jury as confirming that such information would be material to reasonable investors.

Defendants argue that the Individual Defendants relied on the advice of lower-level employees, and that there is no evidence to suggest they intended to deceive investors by failing to disclose the LPM problem sooner. But the emails cited above create a genuine issue of fact as to whether Ahearn, Sohn, Eaglesham, Meyerhoff, Zhu, and Gillette were personally involved in managing the LPM problem. Several emails indicate that Defendants wanted to keep the information from becoming public, and took steps to avoid disclosing it in press releases and conference calls. The fact that lower-level employees may have recommended these practices does not necessarily absolve Defendants of responsibility.¹¹

11 The Court notes Plaintiffs' failure to connect Widmar to the alleged scheme to conceal the existence of the LPM defect from the public. In fact, Widmar did not become CFO of First Solar until early 2011. As such, he cannot be responsible for this alleged fraud.

In light of the evidence set forth above, the Court concludes that Plaintiffs have established genuine issues of material fact regarding whether Defendants had a duty to disclose the manufacturing defect prior to July 2010, and whether such information would have been material to a reasonable investor. This is not simply a case of a company refraining from “bury[ing] the shareholders in an avalanche of trivial information.” In re Convergent Tech. Sec. Litig., 948 F.2d 507, 516 (9th Cir.1991). A jury could conclude that a reasonable investor would consider the existence of the LPM defect as significantly altering the “total mix of information made available.” Basic, 485 U.S. at 231–32, 108 S.Ct. 978 (internal quotation marks omitted). In addition, based on the communications between the Individual Defendants prior to the disclosure of the LPM defect, a reasonable jury could find that Defendants intended to deceive investors by concealing its existence.

2. Concealment.

Plaintiffs claim that First Solar's initial disclosure of the LPM defect in 2Q10 was misleading because it bounded the problem to “less than 4% of the total product manufactured within the period,” which is equivalent to approximately 400,000 modules or 30MW. Doc. 359-1 at 41. The initial disclosure also stated that “[w]e have been working directly with impacted customers to replace the affected modules and these efforts are well underway and, in some cases, complete.” Id. This disclosure was repeated in a substantially similar form in First Solar's quarterly filings until 4Q11. Plaintiffs assert that Ahearn, Gillette, Eaglesham, Sohn, Meyerhoff, and Zhu reviewed and approved this disclosure before it was released. Doc. 363 at 23.

In April 2010, Koralewski was informed via email that “[t]he 415K is the limit we have at this time. It is based on the probability of returns, not necessarily the total LPM[.]” Doc. 364-3 at 78. In June 2010, First Solar discovered that the LPM population could be as large as 1.4 million modules. Doc. 365-9 at 43. In December 2010, Koralewski noted: “We know that there are more LPM out there (warranty worse than 10%) of approximately 500,000 (addition to what is in model)[.]” Id. at 56.

Plaintiffs also claim First Solar misrepresented the status of its remediation program to investors. An internal report, which was sent to Zhu, noted that as of July 22, 2011, over 2,000 claims had not yet been assessed for validity. Doc. 368-2 at 8. On November 3, 2011, Widmar stated that “[w]e have substantially concluded the remediation programs associated with this manufacturing excursion.” Doc. 311 at 59. In an email circulated between Kallenbach, Koralewski, Zhu and Schumaker discussing the upcoming 3Q11 earnings call and agreeing to state that the remediation programs had substantially concluded, Kallenbach responded: “To be crystal clear the operative term is ‘substantially concluded.’ ” Doc. 368-4 at 29. At his deposition, Thomas Kuster testified that at the end of 3Q11 the remediation programs “had not been substantially concluded. There was still work to be done.” Doc. 364-1 at 21. In 4Q11, First Solar added $125.8 million to the LPM Remediation Accrual in its earnings statement. In the related conference call, Ahearn explained that the additional accrual occurred because of a late influx of customer claims at the end of the quarter.¹²

12 Importantly, the Individual Defendants do not dispute that they authorized all the statements made in First Solar's SEC filings and conference calls. Doc. 311 at 47.

Defendants claim that there is no evidence that the analysis underlying Koralewski's calculation limiting the number of LPMs to 4% of the total product manufactured (or 400,000 modules) was incorrect. They point to Koralewski's declaration, which states that he used his best scientific and engineering judgment in arriving at this conclusion. Doc. 324, ¶ 7. In addition, other technical personnel agreed with his analysis. Id., ¶ 14. Koralewski claims that any documents purporting to demonstrate a higher number of LPMs actually “refer to a larger universe of modules than the population of LPMs manufactured during the excursion.” Id., ¶ 18. This was because “we started using a statistical Monte Carlo model to estimate the total number of modules, including good modules, that First Solar would have to rip and replace to complete the remediation effort.” Id. This did not “differentiate LPMs from modules that experienced power loss for other reasons.” Id.

The evidence presented by Plaintiffs and Defendants creates an issue of fact on whether the total population of LPMs was larger than disclosed in July 2010. The emails cited by Plaintiffs do not differentiate between LPMs generally and LPMs produced as a result of the excursion, and they identified a potential population of LPMs much larger than the 400,000 that was later disclosed. Viewing the facts in a light most favorable to Plaintiffs, as the Court must do at the summary judgment stage, the Court concludes that a reasonable jury could find that the 4% disclosure was materially misleading in light of the information known to Defendants.

In addition, the Court finds a question of fact regarding whether Widmar's statement that the remediation claims process had substantially concluded was misleading to investors. Defendants assert that 89% of the claims had been resolved when Widmar made that statement. Doc. 311 at 59. Defendants also claim that the charge related to a large influx of customer claims at the end of the quarter. But evidence provided by Plaintiffs calls this explanation into question. Over 2,000 claims remained outstanding at the end of July 2011, and 4Q11 brought with it a major accrual charge for remediation after Widmar's statement. A reasonable jury could find that Widmar's November 3 statement was materially misleading.

Defendants argue that they reasonably relied on the data and analysis provided by highly trained scientists and engineers and thus lacked the intent to deceive investors when they made statements in earnings releases and conference calls. Generally, Plaintiffs must show scienter with respect to each Individual Defendant. See Apollo, 774 F.3d at 607. Scienter may be imputed to “individual defendants in some situations, for example, where we find that ‘a company's public statements [are] so important and so dramatically false that they would create a strong inference that at least some corporate officials knew of the falsity upon publication.” Id. at 607–08. Here, the 4% disclosure and Widmar's statement were made in connection with earnings releases, with which the Individual Defendants were heavily involved. And the LPM defect was a major issue facing First Solar at the time. In fact, it was considered “the biggest smoking gun” at the company. A reasonable jury could find that Defendants authorized these statements with the intent to mislead investors, or at least acted recklessly in approving such statements. See Daou, 411 F.3d at 1015.

3. Underaccrual.

Plaintiffs submit the declaration of D. Paul Regan, a CPA of more than 40 years, to support their argument that Defendants intentionally underestimated the financial impact of the LPM defect in their financial statements. Doc. 373, ¶¶ 5-6. Regan opines that First Solar violated GAAP in accruing potential warranty claims related to the LPM problem. Id., ¶¶ 42-46. Specifically, based on the information available at the time, First Solar “failed to make appropriate MD&A disclosures concerning the risk that LPM-related warranty estimates were likely to change.” Id., ¶ 45. Regan identifies several other improper accounting procedures and violations of GAAP. Id., ¶¶ 59-70; 75-81. In addition, he notes that PwC did not audit several key disclosures, such as the 4% disclosure in July 2010. Id., ¶ 91. Moreover, “PwC's audits were not designed to obtain evidence such as e-mails among First Solar's employees.” Id., ¶ 92. PwC also required First Solar to provide verifications of facts as well as “numerous representations regarding its financial statements” on which PwC could rely in completing its audits. Id., ¶ 93.

Defendants assert that it is undisputed that Schumaker did not believe that it was reasonably possible that there would be an increase to the LPM Remediation Accrual that would be material to First Solar's financial condition. But a jury could choose not to credit Schumaker's subjective belief given evidence that Defendants may have been ignoring the true scope of the LPM defect. And Defendants' argument that Regan's conclusions should not be substituted for the engineering and commercial judgments of First Solar employees overlooks the fact that Defendants do not specifically challenge any of Regan's findings. These are factual and credibility issues for the jury to resolve.

The evidence cited by Plaintiffs creates a genuine dispute of fact regarding whether First Solar engaged in accounting violations. Regan's analysis calls into question the accounting methodologies used by First Solar after the LPM defect was disclosed. As noted above, there is evidence that Defendants made misrepresentations regarding the status of the remediation programs. Taken together, a reasonable jury could find that First Solar engaged in accounting fraud.

Defendants again assert that they were entitled to rely on the analysis provided by lower-level employees. They assert that First Solar maintained a rigorous disclosure process and that PwC audited and confirmed its financial statements. But Regan opines that PwC relied on misrepresentations made by Defendants regarding the facts underlying their financial statements. “If it is true that defendants withheld material information from their accountants, defendants will not be able to rely on their accountant's advice as proof of good faith.” Provenz v. Miller, 102 F.3d 1478, 1491 (9th Cir.1996). Defendants do not dispute that PwC was not responsible for confirming the validity of several key facts underlying First Solar's justification for certain accounting practices. And there is a question of fact regarding whether First Solar misrepresented facts on which PwC based its audits. From the same evidence cited above, a reasonable jury could find First Solar engaged in accounting to conceal the true scope of the LPM defect from investors.

B. Hot Climate Degradation.

1. Failure to Disclose and Concealment.

Plaintiffs assert that Defendants discovered the hot climate degradation issue in November 2009 when the Director of Product management disputed First Solar's representation that modules would degrade at a rate of 0.7%-0.8% in hot climates. Doc. 365-5 at 30. He warned that “[f]ield data is noisy, leading to inability to draw definitive conclusion on long term degradation.” Doc. 369-1 at 10. In January 2010, an employee noted that the “new degradation rates are higher [than] expected for hot climates.” Id. at 96. In April 2010, Adrianne Kimber, First Solar's Director of Performance and Production, along with a technical team, provided more data and analysis regarding degradation rates and concluded that “total system energy yield in the first 5 years would be less than our current guidance based on -0.7% annual degradation.” Doc. 369-2 at 13. She noted that “there is a 95% chance that the true hot climate PV system degradation is greater than 0.7%/year.” Id. at 20. The team noted that “First Solar should contemplate a change to external guidance for degradation rates of systems and modules installed in hot climates.” Id. at 13. In addition, the team concluded that the data showed modules would “fall [ ] short of customer expectations by year 7.” Id. Koralewski read the report and concluded that “it is a nice piece of work and data based as we can get at this time.” Doc. 369-3 at 58.

On March 17, 2011, Eaglesham was notified that the issue was raised in a staff meeting and “caused quite a bit of excitement about what should/shouldn't be changed in our financial assumptions.” Id. at 65. Afterwards, an email was sent to First Solar employees, including Meyerhoff and Eaglesham, which addressed the hot climate issue and noted that “[u]ntil we receive executive approval to modify our guidance, there is no change to our degradation guidance in any region.” Id. at 70. In late March, a presentation was made to “[o]rient E-Staff to stabilization issue,” which noted that “hotter sites drop faster, more severely.” Doc. 369-4 at 3, 10. Defendants do not dispute that the seven Individual Defendants were members of E-Staff. Meyerhoff, Zhu, and Sohn admitted that they were part of E-Staff in their depositions. Doc. 364-1 at 118, 170, 215.

In order to allegedly conceal the defect from customers, Plaintiffs claim Defendants “de-rated” modules. De-rating is the “process by which First Solar labeled modules with wattage lower than the wattage at which the module tests, to accommodate [the] exponential part of the [hot climate] degradation.” Doc. 363 at 49 n.38. De-rating was the “simplest, fastest solution,” but would result in a large impact to profit margin. Doc. 369-4 at 14. Eaglesham noted that a large financial impact was imminent. Doc. 369-9 at 2. In an email to Sohn and Meyerhoff, Kallenbach listed the options for dealing with the defect: “(1) Change our label from +/-5% to +5/-10%, (2) Change the derate so that no (almost no) modules fall below -10%, (3) Change the derate so that no (almost no) modules fall below -5%, (4) Offer a system level performance warranty in lieu of a module performance warranty, (5) Do nothing.” Id. at 14.

On April 27, 2011, Eaglesham informed E-Staff and the Board of Directors that modules were degrading at a rate of 11% in hot climates. Doc. 364-1 at 240. Mitigation options, including de-rating, were presented, and Eaglesham informed the Board that the impact of de-rating would be $30-$60 million in 2011. Id. at 241. Ultimately, the defect and its financial impact ($37.8 million) were not disclosed until the 4Q11 release and a conference call in February 2012, nearly a year after Eaglesham's presentation.

Defendants argue that this evidence is misleading because the data was based on an assumption that the modules would degrade linearly. In that case, degradation would be worse than 0.8% per year. But, they assert, modules actually did not degrade linearly, they experienced higher degradation during their first few years in operation, and thus no warranty concerns existed. This argument overlooks evidence that Eaglesham later concluded that modules in hot climates were degrading at a rate of 11%, and that this was going to result in a large financial impact to First Solar.

A reasonable jury could find from the evidence that First Solar's representation that its modules degraded at a rate of 0.7%-0.8% in hot climates was misleading. A reasonable jury could also find that Defendants should have disclosed the hot climate problem sooner and that Defendants concealed the problem from customers to avoid disclosure. The evidence indicates that Eaglesham, Sohn, Meyerhoff, and other members of E-Staff knew about the issue and spent several months trying to mitigate its effects. In April 2011, Eaglesham reported his findings to E-Staff, which concluded that the hot climate degradation issue could have a financial impact of up to $60 million. At that point, all Individual Defendants knew the problem was significant.

In addition, Defendants' concealment of the issue from customers evidences a desire to keep the issue from reaching investors, especially after First Solar had already disclosed the LPM problem less than a year prior. Like the LPM problem, a jury could conclude that a reasonable investor would consider the existence of the hot climate problem as significantly altering the total mix of information made available.

The evidence also creates a question of fact about whether Defendants acted with intent to deceive investors. Defendants argue that only Eaglesham knew of the issue and that he did not believe it to be a problem. But Eaglesham presented his findings to the E-Staff, which included several, if not all, of the Individual Defendants. A jury might also doubt that Eaglesham consider the defect to be a non-problem when he knew it could cost First Solar up to $60 million. What is more, a jury could find that Defendants concealed the defect for several months before disclosing it, and considered five alternative mitigation strategies, which included doing nothing. Instead of notifying customers about the problem, First Solar de-rated modules. A reasonable jury could find that Defendants concealed the existence of the hot climate defect with the intent to mislead investors, or at the very least, acted recklessly in failing to disclose the problem to customers and investors.

2. Underaccrual.

Plaintiffs assert that Defendants concealed the existence of the hot climate degradation for several quarters by engaging in accounting violations so that investors could not learn of the problem through SEC filings. Regan, Plaintiffs' expert, concludes that Defendants retroactively applied the $37.8 million accrual for the hot climate defect that related to modules shipped between 3Q09 through 2Q11. Doc. 373, ¶ 89. He also concludes that Defendants violated GAAP by failing to increase its warranty accrual once it quantified the financial impact of the defect in 1Q11, at which point it was both qualitatively and quantitatively material. Id. Instead, the accrual was delayed until 4Q11. Id.

Defendants argue that they sufficiently accounted for the financial impact of their mitigation strategies in their financial statements. But Plaintiffs provide expert testimony that First Solar committed several GAAP violations regarding the hot climate issue, giving rise to a genuine dispute of fact that cannot be resolved by summary judgment.

C. CpW.

Plaintiffs argue that Defendants manipulated CpW by excluding and delaying recognition of costs related to the LPM problem and hot climate defect in the CpW calculation, as well as excluding the effects of de-rating. Lower CpW indicated greater manufacturing efficiency, but is not a GAAP metric. Nonetheless, investors were aware of this figure and it was disclosed in First Solar's SEC filings. Plaintiffs assert that CpW was improperly manipulated by one penny in 3Q10, from $0.86 to $0.85. They also point to the internal investigation that occurred relating to CpW.

Defendants argue that CpW includes costs only for manufacturing modules in the current reporting period. Remediation and warranty costs were associated with modules produced in earlier quarters and had no effect on manufacturing. Thus, when First Solar disclosed the LPM defect, it advised investors in 2Q10 that CpW did not include costs related to the remediation program. Doc. 311 at 64.

In his declaration, Meyerhoff states that “CpW is a metric used throughout the photovoltaic industry [and] is not a GAAP metric and does not have a standard definition.” Doc. 316, ¶ 49. “First Solar publicly defined CpW as a period manufacturing cost—that is, the in-period costs of producing one watt for sale.” Id. In addition, Meyerhoff notes that “this period metric is not intended to include costs related to prior or future periods.” Id., ¶ 50.

Plaintiffs have failed to provide evidence from which a reasonable jury could conclude that CpW should have contained costs related to the LPM or hot climate defects. There is no evidence that First Solar ever included prior or future costs in CpW. Although the two defects were occurring simultaneously with new manufacturing, there is no evidence that new modules were being manufactured with less efficiency due to defects in previously-manufactured modules. As Meyerhoff notes, CpW is simply a metric used to measure the actual costs required to manufacture a single watt of power.

In addition, Plaintiffs' allegations relating to the internal investigation do little to suggest Defendants intended to manipulate CpW. It is undisputed that CpW was an important metric, but Plaintiffs do not connect any wrongdoing to the Individual Defendants. Even assuming Wood, who is not a Defendant, was improperly pressuring employees to lower CpW, Plaintiffs do not dispute that First Solar immediately investigated the allegations and, with advice of counsel, found no wrongdoing. The record is devoid of evidence from which a reasonable jury could find that Defendants improperly manipulated CpW with the intent to mislead investors. Notably, Regan does not offer an opinion regarding CpW. See Doc. 373.

D. Section 10(b) Conclusion.

Plaintiffs have presented sufficient evidence for a reasonable jury to find loss causation as defined in Daou with respect to the five disclosure events other than the announcement of Gillette's departure. Plaintiffs have also presented sufficient evidence for a reasonable jury to conclude that Defendants (1) had a duty to disclose the LPM defect to investors and failed to do so in order to mislead investors, (2) concealed the scope of the LPM defect with the intent to mislead investors, (3) had a duty to disclose the hot climate degradation to investors and failed to do so in order to mislead investors, and (4) engaged in accounting fraud with respect to both defects with intent to mislead investors.

Plaintiffs have failed to create a question of fact regarding their CpW claim and have failed to present evidence connecting Widmar to the alleged failure to disclose the LPM problem prior to July 29, 2010. Defendants' motion for summary judgment will be granted on these issues in addition to the October 25, 2011 disclosure of Gillette's departure, and denied on the remaining § 10(b) issues.¹³

13 Plaintiffs also claim that Defendants' stock trades evidence scienter. Because the Court has already found that Plaintiffs can satisfy the scienter element, it need not analyze this argument.

VI. Section 20(a) Claims.

Plaintiffs allege that the Individual Defendants are liable for the § 10(b) violations as “controlling persons” under § 20(a). “As controlling persons, they would be jointly and severally liable for violations of section 10(b) of the [Exchange Act] and Rule 10b-5.” Apollo, 774 F.3d at 603. “To establish a cause of action under this provision, a plaintiff must first prove a primary violation of...Section 10(b) or Rule 10b-5, and then show that the defendant exercised actual power over the primary violator.” In re NVIDIA Corp. Secs. Litig., 768 F.3d 1046, 1052 (9th Cir.2014). Defendants argue that they cannot be held liable for statements made by lower-level employees. But Plaintiffs have established questions of fact with respect to their § 10(b) claims, and the statements and misrepresentations have been connected to all of the Individual Defendants, unless otherwise noted. Thus, a genuine dispute of fact remains as to whether Defendants are liable under § 20(a).

VII. Plaintiffs' Motion for Summary Judgment.

Plaintiffs seek summary judgment on eighteen of Defendants' twenty affirmative defenses. Doc. 309. At oral argument, Defendants agreed that the Court could strike twelve of these defenses because they “are in fact not affirmative defenses.” Doc. 400 at 89. Defenses 1–5, 7, 10–14, and 16 will be stricken (Doc. 123 at 79-83), and the Court will not grant Plaintiffs' motion on this ground.

With respect to the six remaining defenses, Plaintiffs assert that Defendants can produce no supporting evidence. Doc. 310 at 2. These defenses include: Sixth—Assumption of Risk; Eighth—Failure to Mitigate Damages; Ninth—Proportional Allocation of Fault; Fifteenth—Statute of Limitations; Nineteenth—Release; and Twentieth—Res Judicata. Id. at 5. Defendants confirm that they are not aware of facts supporting these defenses as to the named Plaintiffs, but argue that they should be permitted to reserve these defenses as to absent class members and damages allocation.

The Court will not grant summary judgment on these six affirmative defenses. This case will not end if a verdict is entered for Plaintiffs at trial. A procedure will be required to establish the claims of class members. The Court cannot determine at this stage whether any of the remaining affirmative defenses would be relevant in that process, and therefore will not eliminate the defenses now.¹⁴

14 With respect to the proportional allocation of fault defense, the Court notes that the PSLRA provides that a defendant “against whom a final judgment is entered in a private action shall be liable solely for the portion of the judgment that corresponds to the percentage of responsibility of that covered person[.]” 15 U.S.C. § 78u-4(f)(2)(B)(i). A defendant may be jointly and severally liable “only if the trier of fact specifically determines that such covered person knowingly committed a violation of the securities laws.” Id. § 78u-4(f)(2)(A). The applicability of these provisions must be determined in light of the jury's verdict.

VIII. Other Matters.

A. Request for Judicial Notice.

Defendants ask the Court to take judicial notice of the following facts: (1) the daily closing price of First Solar stock as recorded by Yahoo! Finance; (2) First Solar's Forms 8-K, 10-Q, and 10-K filed with the Securities and Exchange Commission (“SEC”); (3) First Solar press releases contained on its website; and (4) excerpts taken from First Solar earnings conference calls recorded and transcribed by Bloomberg LP. Defendants attached the information as exhibits to their request, and Plaintiffs do not oppose the request.

Rule 201 permits courts to take judicial notice of facts “capable of accurate and ready determination by resort to sources whose accuracy cannot reasonably be questioned.” Fed. R. Ev. 201(b)(2). Defendants seek judicial notice of facts from reliable sources. The Court will take judicial notice and assume the accuracy of the facts set forth in Defendants' request.

B. Motions to Seal.

Defendants also filed a motion to seal six exhibits in support of their motion for summary judgment (Doc. 342) and a motion to seal six exhibits accompanying their reply brief (Doc. 387). Defendants claim these documents contain trade secrets and “technical information and key metrics related to the performance of First Solar modules, their failure rates over time, detailed projections on the expected degradation of First Solar modules over their lifetime, and technical data on First Solar's manufacturing process for modules destined for use in hot climates.” Doc. 342 at 2-3.

Documents may be sealed if the court finds “compelling reasons supported by specific factual findings...outweigh the general history of access and the public policies favoring disclosure.” Pintos v. Pac. Creditors Ass'n, 605 F.3d 665, 678 (9th Cir.2010). The Court finds that the exhibits contain trade secrets and key technical information that could result in unwarranted injury to First Solar if disclosed. The Court will grant Defendants' motions to seal. Plaintiffs do not oppose them.

IT IS ORDERED:

1. Defendants' motion for summary judgment (Doc. 311) is granted in part and denied in part.

2. Plaintiffs' motion for summary judgment (Doc. 309) is denied, but Defendants' affirmative defenses 1–5, 7, 10–14, and 16 are stricken.

3. Defendants' request for judicial notice (Doc. 341) is

granted.

4. Defendants' motions to seal (Docs. 342, 387) are

granted.

5. The Court certifies the loss causation issue discussed above for immediate interlocutory appeal under 28 U.S.C. § 1292(b). The issue certified is this: what is the correct test for loss causation in the Ninth Circuit? Can a plaintiff prove loss causation by showing that the very facts misrepresented or omitted by the defendant were a substantial factor in causing the plaintiff's economic loss, even if the fraud itself was not revealed to the market (Nuveen, 730 F.3d at 1120), or must the market actually learn that the defendant engaged in fraud and react to the fraud itself (Oracle, 627 F.3d at 392)? Within 10 days of the entry of this order, either side may petition the Ninth Circuit to decide this issue on immediate appeal. Id. If neither side files such a petition, the Court will schedule a case management conference to set a schedule for completion of expert discovery and clarification of trial issues, and will set a final pretrial conference, at which a firm trial date will be set. If a petition for immediate appeal is filed within 10 days of this order, the Court will stay this action until the Ninth Circuit decides whether to take the appeal and, if it does, the stay will remain in effect until the appeal is decided.