Carlton v. Cannon

Full title: Dave Carlton, et al., Plaintiffs, v. Fred Cannon, et al., Defendants.

Court: United States District Court, S.D. Texas, Houston Division.

Date published: May 4, 2016

184 F. Supp. 3d 428 (S.D. Tex. 2016)

Opinion

CIVIL ACTION NO. H-15-012

2016-05-04

Dave Carlton, et al., Plaintiffs, v. Fred Cannon, et al., Defendants.

Nicholas Porritt, Adam Marc Apton, Levi & Korsinsky, LLP, Washington, DC, R. Dean Gresham, Gresham PC, Dallas, TX, Laurence M. Rosen, Phillip Kim, The Rosen Law Firm, P.A., New York, NY, James L. Gascoyne, Gascoyne Bullion PC, Sugar Land, TX, for Plaintiffs. David D. Sterling, Baker Botts LLP, Cristina Espinoza Rodriguez, Hogan Lovells US LLP, Noelle M. Reed, Wallis Mizell Hampton, Skadden Arps et al., Houston, TX, Gregory D. Chisholm, Peter A. Spaeth, Alexandra Boudreau, Peter J. Kolovos, Wilmer Cutler Pickering Hale and Dorr LLP, Boston, MA, Michael G. Bongiorno, Wilmer Cutler Pickering Hale Dorr LLP, New York, NY, Jan Nielsen Little, Julia Sun Choe, Steven P. Ragland, Justina Kahn Sessions, Keker Van Nest LLP, San Francisco, CA, for Defendants.

Lee H. Rosenthal, United States District Judge

Nicholas Porritt, Adam Marc Apton, Levi & Korsinsky, LLP, Washington, DC, R. Dean Gresham, Gresham PC, Dallas, TX, Laurence M. Rosen, Phillip Kim, The Rosen Law Firm, P.A., New York, NY, James L. Gascoyne, Gascoyne Bullion PC, Sugar Land, TX, for Plaintiffs.

David D. Sterling, Baker Botts LLP, Cristina Espinoza Rodriguez, Hogan Lovells US LLP, Noelle M. Reed, Wallis Mizell Hampton, Skadden Arps et al., Houston, TX, Gregory D. Chisholm, Peter A. Spaeth, Alexandra Boudreau, Peter J. Kolovos, Wilmer Cutler Pickering Hale and Dorr LLP, Boston, MA, Michael G. Bongiorno, Wilmer Cutler Pickering Hale Dorr LLP, New York, NY, Jan Nielsen Little, Julia Sun Choe, Steven P. Ragland, Justina Kahn Sessions, Keker Van Nest LLP, San Francisco, CA, for Defendants.

MEMORANDUM AND OPINION

Lee H. Rosenthal, United States District Judge

KiOR, Inc., is a once-promising "clean-tech" start-up company that hoped to commercialize technology to turn woodchips into oil. KiOR attracted prominent investors, media attention, and state and federal funding. The manufacturing of its product, and the company, failed. In this putative class action, investors allege that KiOR violated the federal securities laws in the process. The lead plaintiffs, Dave Carlton and Sharon Kegerreis, sued KiOR; its CEO, Fred Cannon; its CFO, John Karnes; and its cofounder and majority shareholder, Vinod Khosla, under the Securities Exchange Act of 1934, 15 U.S.C. §§ 78a –78pp, seeking to represent purchasers of KiOR's stock between June 24, 2011 and March 17, 2014.

KiOR filed for bankruptcy protection after this suit was filed. The court stayed the claims against KiOR and severed the claims against the individual defendants. See 11 U.S.C. § 362(a)(1) ; (Docket Entry No. 84). The plaintiffs filed a third amended complaint against Cannon, Karnes, and Khosla, alleging that Cannon and Karnes violated § 10(b) of the Exchange Act, 15 U.S.C. § 78j(b), and Rule 10b-5, 17 C.F.R. § 240.10b-5 ; and that all three individual defendants violated § 20(a) of the Exchange Act, 15 U.S.C. § 78t(a). The plaintiffs alleged that the individual defendants knowingly made false and misleading statements to investors in KiOR's registration prospectus, quarterly and annual SEC filings, earnings calls, press releases, and investor presentations. The statements concerned the yields and costs at KiOR's demonstration unit in Houston, Texas; the hurdles facing KiOR's production capabilities at its commercial-scale plant in Columbus, Mississippi; and the projected future yields at the Columbus plant.

The individual defendants moved to dismiss. (Docket Entry Nos. 92, 93, 94). Cannon and Karnes argued that the safe-harbor provision in the Private Securities Litigation Reform Act of 1995 ("PSLRA"), Pub. L. No. 104–67, 109 Stat. 737, applies; that the third amended complaint failed to allege the falsity of their statements with particularity; and that the complaint failed to allege facts giving rise to a strong inference of scienter. Khosla argued that he lacked the ability to control either the company's day-to-day operations or the specific misstatements alleged. The plaintiffs filed an omnibus response, (Docket Entry No. 97), and the defendants replied, (Docket Entry Nos. 99, 100, 101). The court heard oral argument. (Docket Entry Nos. 102, 107).

Based on the pleadings, the motions and responses, the applicable law, and counsel's arguments, the court grants Khosla's and Karnes's motions to dismiss. The claims against them are dismissed with prejudice because additional amendment would be futile. The court denies Cannon's motion to dismiss. His statements from November 2012 to January 2014 about how the Columbus plant operated are actionable. See Part IV.A. Some of Cannon's statements cannot be the basis for liability, however. These statements include:

• The statements about the Houston demonstration unit, see Part III;

• The oral statements about future yields, see Part IV.B; and

• The written press-release statements about future yields, to the extent they are forward looking, see Part IV.C.

The reasons for these rulings are explained below. A status and scheduling conference is set for May 25, 2016 at 10:00 a.m. to set a scheduling order for discovery, subsequent motions, and trial.

Table of Contents

I. Background...445

A. The Plaintiffs' Allegations...445

1. The Technology...445

2. The Houston Demonstration Unit...446

3. The Columbus Commercial-Scale Plant...446

B. The Allegations About Information Provided by the Confidential Witnesses... 447

C. Allegations About Information from KiOR's Chief Technology Officer, Paul O'Connor...449

II. The Legal Standards...450

A. Rule 12(b)(6)...450

B. The Section 10(b) and Rule 10b-5 Claims Against Cannon and Karnes...452

1. The Safe-Harbor Provision...452

a. The Statutory Definition...452

b. The Cautionary Statement Must Be Meaningful...454

c. The Cautionary Statement May not Be Immaterial...455

d. The Test for Forward-Looking Oral Statements...455

2. Falsity...456

3. Scienter...458

4. Loss Causation...460

C. The Section 20 Claims Against Cannon, Karnes, and Khosla...461

III. Statements About the Demonstration Unit in Houston, Texas...461

A. The Safe Harbor...462

B. Loss Causation...464

IV. Statements About How the Columbus, Mississippi Plant Operated...465

A. Representations of Then-Current Facts...465

1. The Safe Harbor...467

2. Falsity...468

3. Scienter...475

a. Karnes...476

b. Cannon...482

4. Summary of the Conclusions...484

B. The Oral Statements About Future Yields...485

1. Whether the Projections Were Accompanied by Meaningful Cautions...485

2. Whether the Projections Were Made with "Actual Knowledge"...493

C. The Written Statements About Future Yields...493

V. Control-Person Liability...495

VI. Conclusion...496

I. Background

A. The Plaintiffs' Allegations

1. The Technology

KiOR, a renewable-fuels company incorporated in 2007, sought to commercialize technology to convert nonfood biomass like woodchips into "green" gasoline for cars and other vehicles. (Docket Entry No. 86 at ¶ 76). The technology called for a process to reduce woodchips to a dry, flour-like dust and funnel the dust into a catalytic reactor that "cracks" the woodchips' molecular structure to vaporize and rearrange it. (Id. at ¶ 3). This process would produce crude oil and other byproducts, including light gases, water, and coke. The mixture would enter a separator to remove the catalyst from the oil and related byproducts and transfer them to a product-recovery unit for refining and a hydrotreating unit for processing into gasoline or diesel. The catalyst would be sent to a regenerator for cleaning and recycling. (Id. at ¶ 77).

The reactor, separator, and catalyst regenerator are part of the "biomass fluid-catalytic-cracking process." This is KiOR's core proprietary technology. The "yield"—the fuel this process extracts from the woodchips—is measured in gallons of fuel per one "bone-dry" ton of biomass.

The third amended complaint sets out extensive information from nonparty informants. Only one, Paul O'Connor, is identified by name. Khosla and Paul O'Connor founded KiOR in 2007. Khosla provided the initial financing, owned or controlled more than 70 percent of the voting stock, and appointed his business partner, Samir Kaul, to the board. O'Connor worked with Confidential Witness 6 ("CW 6"), who joined KiOR in 2007 as its senior scientist to develop the technology that KiOR tried to commercialize. O'Connor served on the board from November 2007 until June 2012, and again from March 18, 2014 to August 31, 2014. He was also KiOR's chief technology officer from 2007 to 2009.

2. The Houston Demonstration Unit

KiOR developed a demonstration unit near Houston, Texas to test its technology. Based on the demonstration unit's results, KiOR went public in June 2011. The IPO materials stated that KiOR had achieved conversion yields of 67 gallons of fuel per bone-dry ton and that a commercial-scale plant could consistently achieve this yield at a production cost of less than $1.80 per gallon. The plaintiffs alleged that this representation was false, describing information obtained from CW 6. This information described problems at the demonstration unit and data showing that the unit yields were at most 40 gallons and cost at least $6 per gallon. The plaintiffs alleged that even after CW 6 told Cannon and Karnes about these concerns, they continued to misrepresent the figures in public statements.

3. The Columbus Commercial-Scale Plant

KiOR built its first commercial-scale production plant in Columbus, Mississippi, in a building that once housed a paper mill. The plant was mechanically complete by April 2012 and fully commissioned for operations by September of that year. Soon after the plant began operating, KiOR's officers informed investors that there were some "start-up" issues. Relying on statements from confidential witnesses, all former KiOR employees, the plaintiffs alleged that these issues were serious and fundamental design flaws, not temporary "start-up" problems. The flaws included undersized equipment, tar buildup, unexpected production shutdowns, and slow-moving conveyor systems. (Docket Entry No. 86 at ¶ 112). Additional problems included design flaws in the vapor-recovery unit, leaks, a city-wide blackout, instrument failures, and operating variability. (Id . at ¶¶ 40–59). The plaintiffs alleged that confidential witnesses and O'Connor told Cannon and Karnes about the severity of these problems, but that they deliberately or recklessly misrepresented the degree to which the problems resulted from fundamental design and structural defects as opposed to temporary start-up glitches. The plaintiffs alleged that from September 2012, when operations began, throughout 2013, Cannon and Karnes continually reassured investors that the plant problems were start-up glitches and that the core technology was fundamentally sound. Cannon and Karnes also made specific predictions to investors about future yields, which the plant never achieved. The plaintiffs alleged that Cannon and Karnes knew that the Columbus plant suffered from serious structural deficiencies, making their predictions deliberately, or at least recklessly, misleading.

In 2013, Cannon and Karnes steadily revised their predictions downward. Public disclosures about yields caused the stock to drop. Karnes resigned as CFO in December 2013. The total 2013 yield was 894,000 gallons—well under the 3 to 5 million gallons Cannon and Karnes had predicted at the beginning of the year. (Compare id. at ¶ 147, with id. at ¶ 117).

In January 2014, KiOR announced that it would temporarily shut down the Columbus plant during the first quarter of 2014. Cannon told investors that the plant faced "structural design bottlenecks and reliability issues" that limited the amount of wood that could be introduced into the plant, and he outlined plans to improve the plant's yield, efficiency, and reliability. (Id. at ¶ 150). In March 2014, KiOR shut the Columbus plant down for good. Cannon reported that the plant was unable to achieve its yield targets because of "structural bottlenecks, reliability and mechanical issues, and catalyst performance." (Id. at ¶ 158).

B. The Allegations About Information Provided by the Confidential Witnesses

The third amended complaint relied on information from nine confidential witnesses. The alleged roles and information are described below.

• CW 1 was a commissioning and start-up consultant who worked at the Columbus plant from June 2012 to February 2013. CW 1 had previously worked in the petroleum, chemical, and power industries as a project manager and engineer. CW 1 reported to CW 2, whose role and background are described below. CW 1 had personal knowledge from working at the Columbus plant that the vapor-recovery unit suffered from "significant design flaws" that caused "production shortfalls." CW 1 concluded that the unit needed to be "redesigned before being able to run smoothly." (Docket Entry No. 86 at ¶¶ 40–41).

• CW 2 was an operations superintendent responsible for managing the Columbus plant's day-to-day operations between June 2012 and February 2013. CW 2 had previously worked in project planning and plant commissioning at several refineries. CW 2 reported to CW 4. Based in part on CW 2's observations and experience, the plaintiffs alleged that the plant had "operational problems" stemming from "design issues," including undersized equipment, a fractionating system prone to excess tar buildup, and a woodchip-feeding system that could not supply operations at the necessary rate. CW 2 attended daily meetings about plant operations and was in frequent contact with engineers from KiOR's Texas office about the plant. CW 2 reported that "the whole damn thing was a design issue. It was hastily put together and they were trying to make it run." CW 2 asserted that "within just a few days of operations," the problems were "apparent to everyone." (Id . at ¶¶ 42–43).

• CW 3 worked at the Columbus plant as a millwright in the summer of 2013. CW 3 stated that the plant did not operate from July to August, and there were significant production

problems when operations finally began. CW 3 saw and reported excessive amounts of a tar-like substance in fuel-product tanks. CW 3 also saw and reported persistent problems with the conveyer system that fed biomass into the plant's operations. (Id . at ¶ 44).

• CW 4 was KiOR's operations manager at the Columbus plant from September 2011 to February 2013. CW 4 reported to Greg Neve, the plant manager. CW 4 attended daily meetings with managers and engineers, who reported the problems to Cannon. CW 4 saw equipment failures affecting the wood-chipping unit, the biomass fluid-catalytic-cracking system, and the vapor-recovery unit, including undersized equipment. CW 4 stated that senior management was "aware of these mechanical and design problems" because they were "discussed constantly." (Id . at ¶¶ 45–46).

• CW 5 worked for Austin Industrial, KiOR's maintenance contractor. CW 5 provided maintenance services at the Columbus plant for two weeks in September 2013. CW 5 recalled spending significant time repairing the wood-chipping and conveyer-feeding systems. One repair to the conveyer-feeding system required multiple maintenance contractors to work three, twelve-hour shifts. According to CW 5, senior management was "definitely aware" of the problems related to the wood-chipping system because the engineers and managers "all knew about the ongoing problems." (Id. at ¶ 47).

• CW 6, KiOR's senior scientist beginning in 2007, worked with Paul O'Connor to develop KiOR's technology. In 2010, when Khosla wanted to take the company public, CW 6 told him that the technology "was not yet meeting expectations" and "was not yet fully developed." CW 6 described the information provided in KiOR's 2011 registration prospectus as inaccurate. The prospectus reported a yield of 67 gallons per bone-dry ton at a cost of $1.80 per gallon. CW 6 "was aware of" data showing a yield of 40 gallons per bone-dry ton at a cost of $6 to $7 per gallon. CW 6 personally knew this information because he worked full time at KiOR's demonstration unit in Houston.

CW 6 complained to Cannon, Karnes, and other managers about the inaccurate statements, both during in-person meetings and by email. CW 6 resigned from KiOR's executive-management team in the latter part of 2011 because of the misrepresentations about the strength of KiOR's technology. He continued to serve as a consultant, however. In that capacity, he complained in executive-management meetings to Cannon, Karnes, and others about repeated misrepresentations. CW 6 met with KiOR's COO, William Coates, in August or September 2011 and told Coates that KiOR's technology could not produce the projected yields at all, much less at the projected costs. After this meeting, according to CW 6, Coates told Cannon about the conversation, and Coates was either fired or resigned as a result.

CW 6 served as a consultant throughout 2012 and 2013. In mid-2013, a KiOR staff engineer, Charlie Zhang, told CW 6 that the Columbus plant was producing at most 22 gallons per bone-dry ton. Zhang also told CW 6 that costs per gallon were much higher than the costs KiOR reported. CW 6 arranged to meet with Cannon and Christopher Artzer, a KiOR vice-president, in August or September 2013. CW 6 sent them a letter before

the meeting stating his fear that the company was "disseminating misinformation." The plaintiffs alleged that at the meeting, CW 6 again told Cannon and Artzer about his concerns and promised to raise them with the board of directors. In response, Cannon asked CW 6 to sign a separation agreement, but CW 6 refused. CW 6's consulting contract expired in October 2013 and was not renewed. (Id . at ¶¶ 48–54).

• CW 7 was a construction manager at the Columbus plant from May 2011 to August 2014, responsible for construction work, improvement projects, and mechanical issues. CW 7 reported to the plant manager, Michael O'Keefe. CW 7 specialized in fluid-catalytic-cracking operations. The plaintiffs alleged that CW 7 provided information about the plant's inability to produce the projected yields because of chemical, catalyst, and wood-ratio problems in the catalytic-conversion process. CW 7 alleged that Cannon's public statements overstated the plant's capabilities. According to CW 7, the yield was 29 gallons per bone-dry ton during the time that Cannon and the company publicly represented a 60-gallon yield. CW 7 participated in a May 2013 52-day production run, which made it "evident" that the plant could not meet the projected yields. CW 7 was "uncomfortable with the information KiOR was reporting to the public" because there was not a "snowball's chance" of meeting the projections in light of the production issues. (Id. ¶ 55).

• CW 8 was KIOR's senior vice-president of operations beginning in September 2011. Although KiOR stated that CW 8's employment ended in March 2013, it in fact ended in July 2012. CW 8 left the company because, even though he was an officer, he could not access data about the Columbus plant's performance. CW 8 knew about the plant's low yields and found Cannon's optimism "startling" because the projections "stretched credibility." (Id . ¶¶ 56–58).

• CW 9 worked as a refinery operator from March 2011 to August 2014. According to CW 9, the Columbus plant's problems were obvious from the Houston demonstration unit but were never resolved. CW 9 stated that "[KiOR] never did anything to change it. They couldn't figure it out...so they decided to ignore [it] and hope it wouldn't happen" later at the Columbus plant. (Id . at ¶ 59).

C. Allegations About Information from KiOR's Chief Technology Officer, Paul O'Connor

The third amended complaint also alleged information obtained from Paul O'Connor, KiOR's cofounder, chief technology officer from 2007 to 2009, and board member from November 2007 to June 2012 and again from March 18, 2014 to August 31, 2014. (Id . at ¶¶ 60–75). The third amended complaint alleged that O'Connor became concerned about KiOR's operations in late 2011. He was "shocked" when he reviewed data showing that KiOR had "hardly [made] any progress... since the end of 2009." After O'Connor told Cannon about his concerns and conducted a "technology review" of KiOR's operations, O'Connor concluded that: (1) the low yields resulted in part from "lower than expected oil recovery from the oil-water separation"; (2) the yields had not improved significantly since 2009; and (3) although it was "possible" to reach the yield target of 67 gallons per bone-dry ton, that would "require a drastically different approach[ ]." (Id. at ¶ 64). In March 2012, O'Connor conducted another review, which again showed "very low" yields. (Id. at ¶ 71). O'Connor gave the report to Khosla and Khosla's appointed board member, Samir Kaul, who "always" relayed O'Connor's messages to Cannon. Cannon's response was that O'Connor was "being too pessimistic," and Cannon took no action.

O'Connor recalled that, at an April or May 2012 board meeting, KiOR's research-and-development director admitted that the company "should not expect to reach the [goal of] 67 [gallons per bone-dry ton]." Instead, the Columbus plant could expect yields between 30 to 40 gallons per bone-dry ton. (Id . ¶ 65).

O'Connor wrote Cannon a letter before a May 2013 annual shareholder meeting, expressing concerns about KiOR's technology. Cannon told O'Connor that he was being "too negative" and that KiOR had made "tremendous progress." In September 2013, when O'Connor questioned Kaul about production levels, Kaul responded that he could not reveal the information because O'Connor was no longer a director. (Id. at ¶ 74). At first, O'Connor, like Cannon, thought that the Columbus plant's problems were "related to normal ‘start-up’ issues." He changed his view after a November 2013 audit made it clear that yields would be much lower than expected for an extended period. (Id . at ¶ 66).

In January 2014, KiOR sought O'Connor's help for a technical-expert review of the Columbus plant. O'Connor concluded that "the main problems at Columbus [were] already discernable in the [Houston demonstration unit] and [were] therefore structural and not ‘just’ operational issues." (Id . at ¶ 67). The next month, O'Connor told the board that KiOR needed both "a clear change in technology strategy" and in "leadership style" to solve the Columbus production problems. According to O'Connor, the Houston demonstration unit's performance justified only low yield predictions and showed that the Columbus plant was "never suited to produce the yields and/or amounts forecasted." (Id . at ¶ 69).

Deciding the motions to dismiss requires applying the legal standards under Rule 12(b)(6) of the Federal Rules of Civil Procedure and under the Exchange Act, as amended by the PSLRA, to the allegations in the third amended complaint. The legal standards are set out below.

II. The Legal Standards

A.Rule 12(b)(6)

A pleading is deficient and may be dismissed under Rule 12(b)(6) if a plaintiff fails "to state a claim upon which relief can be granted." FED. R. CIV. P . 12(b)(6). Rule 12(b)(6) is read in conjunction with Rule 8(a), which requires "a short and plain statement of the claim showing that the pleader is entitled to relief." FED. R. CIV. P . 8(a)(2). A complaint must contain "enough facts to state a claim to relief that is plausible on its face." Bell Atlantic Corp. v. Twombly , 550 U.S. 544, 555, 127 S.Ct. 1955, 167 L.Ed.2d 929 (2007) ; Ashcroft v. Iqbal , 556 U.S. 662, 129 S.Ct. 1937, 173 L.Ed.2d 868 (2009). Rule 8"does not require ‘detailed factual allegations,’ but it demands more than an unadorned, the-defendant-unlawfully-harmed-me accusation." Iqbal , 556 U.S. at 678, 129 S.Ct. 1937 (citing Twombly , 550 U.S. at 555, 127 S.Ct. 1955 ). "A claim has facial plausibility when the plaintiff pleads factual content that allows the court to draw the reasonable inference that the defendant is liable for the misconduct alleged." Id. (citing Twombly , 550 U.S. at 556, 127 S.Ct. 1955 ). "The plausibility standard is not akin to a ‘probability requirement,’ but it asks for more than a sheer possibility that a defendant has acted unlawfully." Id. (citing Twombly , 550 U.S. at 556, 127 S.Ct. 1955 ). To withstand a Rule 12(b)(6) motion, a "complaint must allege ‘more than labels and conclusions,’ " and " ‘a formulaic recitation of the elements of a cause of action will not do.’ " Norris v. Hearst Trust , 500 F.3d 454, 464 (5th Cir.2007) (quoting Twombly , 550 U.S. at 555, 127 S.Ct. 1955 ). "Nor does a complaint suffice if it tenders ‘naked assertion[s]’ devoid of ‘further factual enhancement.’ " Iqbal , 556 U.S. at 678, 129 S.Ct. 1937 (alteration in original) (quoting Twombly , 550 U.S. at 557, 127 S.Ct. 1955 ). "To survive a Rule 12(b)(6) motion to dismiss, a complaint ‘does not need detailed factual allegations,’ but must provide the plaintiff's grounds for entitlement to relief—including factual allegations that when assumed to be true ‘raise a right to relief above the speculative level.’ " Cuvillier v. Taylor , 503 F.3d 397, 401 (5th Cir.2007) (footnote omitted) (quoting Twombly , 550 U.S. at 555, 127 S.Ct. 1955 ). "Conversely, ‘when the allegations in a complaint, however true, could not raise a claim of entitlement to relief, this basic deficiency should...be exposed at the point of minimum expenditure of time and money by the parties and the court.’ " Id. (quoting Twombly , 550 U.S. at 558, 127 S.Ct. 1955 ).

When a plaintiff's complaint fails to state a claim, the court should generally allow the plaintiff to amend the complaint under Rule 15(a) before dismissing the action with prejudice, unless it is clear that doing so would be futile. See Great Plains Trust Co. v. Morgan Stanley Dean Witter & Co. , 313 F.3d 305, 329 (5th Cir.2002) ("[D]istrict courts often afford plaintiffs at least one opportunity to cure pleading deficiencies before dismissing a case, unless it is clear that the defects are incurable or the plaintiffs advise the court that they are unwilling or unable to amend in a manner that will avoid dismissal."). However, a plaintiff should be denied leave to amend a complaint if the court determines that the "proposed amendment clearly is frivolous, advancing a claim or defense that is legally insufficient on its face." 6 WRIGHT & MILLER § 1487 ; see also Ayers v. Johnson , 247 Fed.Appx. 534, 535 (5th Cir.2007) (" ‘[A] district court acts within its discretion when dismissing a motion to amend that is frivolous or futile.’ " (quoting Martin's Herend Imports, Inc. v. Diamond & Gem Trading U.S. of Am. Co. , 195 F.3d 765, 771 (5th Cir.1999) )).

In considering a Rule 12(b)(6) motion to dismiss, a court limits itself to the contents of the pleadings, with an exception. In Collins v. Morgan Stanley Dean Witter , 224 F.3d 496, 498–99 (5th Cir.2000), the Fifth Circuit approved the district court's consideration of documents the defendant attached to a motion to dismiss. The Fifth Circuit made it clear that "such consideration is limited to documents that are referred to in the plaintiff's complaint and are central to the plaintiff's claim." Scanlan v. Tex. A & M Univ. , 343 F.3d 533, 536 (5th Cir.2003) (citing Collins , 224 F.3d at 498–99 ). Other courts approve the same practice. See Venture Assocs. Corp. v. Zenith Data Sys. Corp. , 987 F.2d 429, 431 (7th Cir.1993) ("Documents that a defendant attaches to a motion to dismiss are considered part of the pleadings if they are referred to in the plaintiff's complaint and are central to her claim." (citations omitted)); see also Field v. Trump , 850 F.2d 938, 949 (2d Cir.1988) (citation omitted); Branch v. Tunnell , 14 F.3d 449, 453–54 (9th Cir.1994), overruled on other grounds by Galbraith v. County of Santa Clara , 307 F.3d 1119 (9th Cir.2002).

Attached to the motions to dismiss are the following documents, which the plaintiffs referred to in the third amended complaint and are central to their claims: KiOR's SEC filings, including its registration prospectus and Forms 8-K, 10-Q, and 10-K; earnings-call transcripts; charts summarizing the SEC filings; and a redline showing changes made to the text of the filings. (Docket Entry No. 94, Ex. 2; Docket Entry No. 99, Ex. 1). The case law permits, and the plaintiffs do not object to, the court considering these documents in deciding the motion to dismiss. Collins , 224 F.3d at 498–99 ; In re Sec. Litig. BMC Software , 183 F.Supp.2d 860, 884 (S.D.Tex.2001) (chart "included only for the court's convenience as a summary of information included in SEC filings" can be considered on a motion to dismiss (citing FED. R. EVID . 1006 )).

B. The Section 10(b) and Rule 10b-5 Claims Against Cannon and Karnes

"The elements of a private securities fraud claim based on Section 10(b) and Rule 10b-5 are (1) a material misrepresentation (or omission), (2) scienter, i.e., a wrongful state of mind, (3) a connection with the purchase or sale of a security, (4) reliance, often referred to in cases involving public securities markets (fraud-on-the-market cases) as transaction causation; (5) economic loss; and (6) loss causation, i.e., a causal connection between the material misrepresentation and the loss." Owens v. Jastrow , 789 F.3d 529, 535 (5th Cir.2015) (quotation marks omitted). Cannon and Karnes moved to dismiss the claims on three grounds: (1) their forward-looking statements are protected under the PSLRA's safe-harbor provision, (2) the third amended complaint fails to allege the falsity of their statements with the requisite particularity, and (3) the third amended complaint does not give rise to a strong inference of scienter. Cannon and Karnes also argued, in the alternative, that the third amended complaint failed to allege a corrective disclosure and therefore failed to allege loss causation as to some of the statements.

1. The Safe-Harbor Provision

a. The Statutory Definition

The PSLRA protects certain "forward-looking statements." A forward-looking statement includes:

(A) a statement containing a projection of revenues, income (including income loss), earnings (including earnings loss) per share, capital expenditures, dividends, capital structure, or other financial items;

(B) a statement of the plans and objectives of management for future operations, including plans or objectives relating to the products or services of the issuer;

(C) a statement of future economic performance, including any such statement contained in a discussion and analysis of financial condition by the management or in the results of operations included pursuant to the rules and regulations of the Commission;

(D) any statement of the assumptions underlying or relating to any statement described in subparagraph (A), (B), or (C); [or]

(E) any report issued by an outside reviewer retained by an issuer, to the extent that the report assesses a forward-looking statement made by the issuer[.]

15 U.S.C. § 78u–5(i)(1). A defendant "shall not be liable with respect to any forward-looking statement, whether written or oral, if":

(A) the forward-looking statement is—

(i) identified as a forward-looking statement, and is accompanied by meaningful cautionary statements identifying important factors that could cause actual results to differ materially from those in the forward-looking statement; or

(ii) immaterial; or

(B) the plaintiff fails to prove that the forward-looking statement—

(i) if made by a natural person, was made with actual knowledge by that person that the statement was false or misleading;

Id. § 78u–5(c)(1). The Fifth Circuit has described the test for applying the safe-harbor provision as "two independent prongs: one focusing on the defendant's cautionary statements and the other on the defendant's state of mind." Southland Sec. v. I NS pire Ins. Sols. , 365 F.3d 353, 371 (5th Cir.2004). Under the first prong, the safe harbor protects a forward-looking statement if it is either (1) identified as forward looking and accompanied by "meaningful cautionary statements" or (2) immaterial. Id. Under the second prong, the safe harbor protects a forward-looking statement if the plaintiff fails to prove that the statement was made with "actual knowledge" that it was false or misleading. Id.

The plaintiffs argue that as long as they sufficiently allege facts showing actual knowledge that the statements were false or misleading, the safe harbor cannot apply as a matter of law. (Docket Entry No. 97 at p. 18). But the PSLRA's safe harbor is disjunctive. See 15 U.S.C. § 78u–5(c)(1) (using the word "or"). Fifth Circuit case law makes clear that the safe harbor applies to a statement that is identified as forward looking and accompanied by meaningful cautionary language, that is immaterial, or that was not made with actual knowledge that it was false or misleading. Lormand v. US Unwired, Inc. , 565 F.3d 228, 243 (5th Cir.2009) (a three-pronged, disjunctive inquiry); Southland , 365 F.3d at 371 ("two independent prongs").¹ Even if the plaintiffs show actual knowledge, the safe harbor may still apply if the statement is either immaterial or is identified as forward looking and is accompanied by meaningful cautionary language. Southland , 365 F.3d at 371 ; see also In re BP p.l.c. Sec. Litig. , 843 F.Supp.2d 712, 777 (S.D.Tex.2012). Other circuits are consistent.² b. The Cautionary Statement Must Be Meaningful

1 Although Lormand framed the test in the disjunctive, the court concluded that the safe harbor did not apply to the alleged misrepresentations because "the plaintiff adequately allege[d] that the defendants actually knew that their statements were misleading at the time they were made."565 F.3d at 244. To the extent Lormand conflicts with Southland 's holding that the safe-harbor provision consists of two "independent prongs," the court follows Southland under the prior-panel-precedent rule. See Smith v. GTE , 236 F.3d 1292, 1300 n. 8 (5th Cir.2001) ("[T]he holding of the first panel to address an issue is the law of this Circuit, thereby binding all subsequent panels unless and until the first panel's holding is overruled by the Court sitting en banc or by the Supreme Court.").

2 See, e.g. , Slayton v. Am. Express Co. , 604 F.3d 758, 766 (2d Cir.2010) ("The safe harbor is written in the disjunctive; that is, a defendant is not liable if the forward-looking statement is identified and accompanied by meaningful cautionary language or is immaterial or the plaintiff fails to prove that it was made with actual knowledge that it was false or misleading."); In re Aetna, Inc. Sec. Litig. , 617 F.3d 272, 278–79 (3d Cir.2010) ("[T]he safe harbor applies to statements that are forward-looking as defined by the statute provided that they are (1) identified as such, and accompanied by meaningful cautionary statements; or (2) immaterial; or (3) made without actual knowledge that the statement was false or misleading."); In re Cutera Sec. Litig. , 610 F.3d 1103, 1112 (9th Cir.2010) ("[T]he investors argue for a conjunctive reading of the safe harbor provision, under which a sufficiently strong inference of actual knowledge would overcome a claim of safe harbor protection even for statements identified as forward-looking and accompanied by meaningful cautionary language. The difficulty with this approach is that it ignores the plain language of the statute, which is written in the disjunctive as to each subpart."); Edward J. Goodman Life Income Trust v. Jabil Circuit, Inc. , 594 F.3d 783, 795 (11th Cir.2010) ("The statute offers several ways for a defendant to avoid liability, all written in the disjunctive....First, a defendant may avoid liability by showing that his statement was issued with meaningful cautionary language. Or, the defendant can show that the statement was simply immaterial. As a third alternative, the defendant can avail himself of the safe harbor if the plaintiff fails to prove that the statement was made with actual knowledge that it was false. Any one of these suffices for the defendant; a top-to-bottom reading of the statute shows that the plaintiff's inability to show knowledge of falsity is only relevant if the defendant is unable to produce meaningful cautionary statements or evidence of immateriality." (citations omitted)); Ind. State Dist. Council & Hod Carriers Pension & Welfare Fund v. Omnicare, Inc. , 583 F.3d 935, 943 (6th Cir.2009) (the safe-harbor provision "is overcome only if the statement was material; if defendants had actual knowledge that it was false or misleading; and if the statement was not identified as ‘forward-looking’ or lacked meaningful cautionary statements." (quotation marks omitted)).

A cautionary statement must "identify important factors that could cause actual results to differ materially from those in the forward-looking statement." 15 U.S.C. § 78u–5(c)(1)(A)(i). This standard is "somewhat ambiguous given the variety of possible factual circumstances that could arise." In re Harman Int'l Indus., Inc. , 791 F.3d 90, 101 (D.C.Cir.2015) (citations omitted). The question is how to separate those disclosures that provide investors with "meaningful" and "important" information from those that are either too vague for an investor to find relevant or too specific to expect a reasonable issuer to identify as important. The case law is clear that while " ‘boilerplate’ warnings won't do; cautions must be tailored to the risks that accompany the particular projections," the cautionary-statement requirement does not demand prescience. Asher v. Baxter Int'l Inc. , 377 F.3d 727, 732 (7th Cir.2004) (collecting cases). "[T]he cautions need not identify what actually goes wrong and causes the projections to be inaccurate." Id. This tension reflects a safe-harbor provision born out of "a compromise between legislators who did not want any safe harbor...and those who wanted a safe harbor...that did not require any cautionary statements but just required the projection to have a reasonable basis." Id.

A "meaningful" cautionary statement "calls for ‘substantive’ company-specific warnings based on a realistic description of the risks applicable to the particular circumstances, not merely a boilerplate litany of generally applicable risk factors." Southland , 365 F.3d at 372. A "generic and formulaic" cautionary statement that is repeated "only with slight variations" and used "in conjunction with each alleged misrepresentation" is not meaningful. Lormand , 565 F.3d at 245 (emphasis in original). Nor are "[s]tatements along the lines of ‘all businesses are risky’ or ‘the future lies ahead,’ " which at bottom "come to nothing other than caveat emptor." Asher , 377 F.3d at 733 ; see also Lormand , 565 F.3d at 244 (the following disclaimer was not meaningfully cautionary: "[The statements are] not guarantees of future performance...and involve known and unknown risks and other factors that could cause actual results to be materially different from any future results expressed or implied by them."); Plotkin v. IP Axess, Inc. , 407 F.3d 690, 694 (5th Cir.2005) (the following disclaimer was not meaningfully cautionary: "These forward-looking statements involve numerous risks, uncertainties and assumptions, and actual results could differ materially from anticipated results.").

A cautionary statement must be tailored "to a particular company's status at a particular time." In re Harman , 791 F.3d at 101. The PSLRA requires a fit between the disclaimer and the challenged forward-looking statement. The disclaimer must warn of what "could cause actual results to differ materially from those in the forward-looking statement," 15 U.S.C. § 78u–5(c)(1)(A)(i), to be useful and relevant to an investor weighing the projection with other factors informing purchase decisions. See In re Harman , 791 F.3d at 102. Even repeated risk disclosures may not be relevant, useful, or otherwise tailored to the challenged statement. See id. ; Institutional Inv'rs Grp. v. Avaya, Inc. , 564 F.3d 242, 256 (3d Cir.2009). The investor must be "warned of risks of a significance similar to that actually realized" so that she "is sufficiently on notice of the danger of the investment to make an intelligent decision about it according to her own preferences for risk and reward." Harris v. Ivax Corp. , 182 F.3d 799, 807 (11th Cir.1999) (quotation marks omitted). That said, while "investors would like to have...a full disclosure of the assumptions and calculations behind the projections[,]...this is not a sensible requirement. Many of the assumptions and calculations would be more useful to a firm's rivals than to its investors." Asher , 377 F.3d at 733. "[D]isclosing assumptions, methods, or confidence intervals" underlying a projection is not required. Id. at 734. "The PSLRA does not require the most helpful caution;... it is enough to point to the principal contingencies that could cause actual results to depart from the projection." Id.

In keeping with the statutory and case-law emphasis on tailored disclosures, "[e]ach statement that benefits from the safe harbor must be addressed individually." Lormand , 565 F.3d at 245. When a statement is forward looking only in part, the "mixed present/future statement is not entitled to the safe harbor with respect to the part of the statement that refers to the present." Spitzberg v. Houston Am. Energy Corp. , 758 F.3d 676, 691 (5th Cir.2014) (quoting Makor Issues & Rights, Ltd. v. Tellabs Inc. ("Tellabs II "), 513 F.3d 702, 705 (7th Cir.2008) ).

c. The Cautionary Statement May not Be Immaterial

A statement "of the vague and optimistic type... cannot support a securities fraud action [because it] contain[s] no concrete factual or material misrepresentation." Southland , 365 F.3d at 372 (quotation marks omitted). "[G]eneralized positive characterization[s] [are] not actionable under the securities laws." Rosenzweig v. Azurix Corp. , 332 F.3d 854, 870 (5th Cir.2003). "Allegations that amount to little more than corporate ‘cheerleading’ are puffery, projections of future performance not worded as guarantees, and are not actionable under federal securities law because no reasonable investor would consider such statements material and because investors and analysts are too sophisticated to rely on vague expressions of optimism rather than specific facts." In re BP , 843 F.Supp.2d at 748 (citing Krim v. BancTexas Grp., Inc. , 989 F.2d 1435, 1446 (5th Cir.1993) ). "[G]eneralized positive statements about a company's progress are not a basis for liability." Nathenson v. Zonagen Inc. , 267 F.3d 400, 419 (5th Cir.2001) ; see also Omnicare , 583 F.3d at 944 ("[Puffery is the] kind of rosy affirmation[s] commonly heard from corporate managers and numbingly familiar to the marketplace—loosely optimistic statements that are so vague, so lacking in specificity...that no reasonable investor could find them important.").

d. The Test for Forward-Looking Oral Statements

The PSLRA states that:

In the case of an oral forward-looking statement..., the requirement set forth

in [§ 78u–5(c)(1)(A) ] shall be deemed to be satisfied–

(A) if the oral forward-looking statement is accompanied by a cautionary statement—

(i) that the particular oral statement is a forward-looking statement; and

(ii) that the actual results might differ materially from those projected in the forward-looking statement; and

(B) if—

(i) the oral forward-looking statement is accompanied by an oral statement that additional information concerning factors that could cause actual results to materially differ from those in the forward-looking statement is contained in a readily available written document, or portion thereof;

(ii) the accompanying oral statement referred to in clause (i) identifies the document, or portion thereof, that contains the additional information about those factors relating to the forward-looking statement; and

(iii) the information contained in that written document is a cautionary statement that satisfies the standard established in paragraph (1)(A).

15 U.S.C. § 78u–5(c)(2).

An "[o]ral statement[ ] can qualify for the safe harbor if (i) the statement is accompanied by a cautionary statement that the ‘particular’ oral statement is forward-looking and that actual results could differ materially (essentially a formality as to the form of the statement); (ii) the statement is accompanied by an oral statement that additional information that could cause actual results to differ materially is contained in a readily-available written document; (iii) the statement identifies the document or portion thereof containing the additional information; and (iv) the identified document itself contains appropriate cautionary language." Southland , 365 F.3d at 372 (citing 15 U.S.C. § 78u–5(c)(2) ). "Readily available written documents for this purpose include documents filed with the SEC or generally disseminated." Id. (citing 15 U.S.C. § 78u–5(c)(3) ).

2. Falsity

Rule 9(b) requires a complaint to "state with particularity the circumstances constituting the fraud." FED. R. CIV. P. 9(b). "Put simply, Rule 9(b) requires ‘the who, what, when, where, and how’ to be laid out." Benchmark Elecs., Inc. v. J.M Huber Corp. , 343 F.3d 719, 724 (5th Cir.2003) ; see also Carroll v. Fort James Corp. , 470 F.3d 1171, 1174 (5th Cir.2006) ("In cases concerning fraudulent misrepresentation and omission of facts, Rule 9(b) typically requires the claimant to plead the type of facts omitted, the place in which the omissions should have appeared, and the way in which the omitted facts made the representations misleading." (quotation marks omitted)). The Fifth Circuit applies " Rule 9(b) to fraud complaints with ‘bite’ and ‘without apology,’ " while recognizing "that Rule 9(b) supplements but does not supplant Rule 8(a)'s notice pleading." United States ex rel. Grubbs v. Kanneganti , 565 F.3d 180, 185–86 (5th Cir.2009) (footnote omitted). Importantly, " Rule 9(b) does not reflect a subscription to fact pleading and requires only simple, concise, and direct allegations of the circumstances constituting fraud, which after Twombly must make relief plausible, not merely conceivable, when taken as true." Id. at 186 (footnote omitted) (quotation marks omitted).

"Securities fraud claims brought by private litigants are also subject to the pleading requirements imposed by the [PSLRA]." Owens , 789 F.3d at 535. A complaint alleging a material misstatement or omission must "specify each statement alleged to have been misleading, the reason or reasons why the statement is misleading, and, if an allegation regarding the statement or omission is made on information and belief, the complaint shall state with particularity all facts on which that belief is formed." 15 U.S.C. § 78u–4(b)(1)(B). "At a minimum, the PSLRA pleading standard incorporates the... requirements of Rule 9(b)." Owens , 789 F.3d at 535 ; see also Cent. Laborers' Pension Fund v. Integrated Elec. Servs. Inc. , 497 F.3d 546, 550 (5th Cir.2007) ("The PSLRA appears to comport with [the Fifth Circuit's] relatively strict interpretation of Rule 9(b), which requires a plaintiff to specify the statements contended to be fraudulent, identify the speaker, state when and where the statements were made, and explain why the statements were fraudulent." (quotation marks omitted)); In re BP , 843 F.Supp.2d at 746 ("[A] plaintiff must...(1) specify each statement alleged to have been misleading; (2) identify the speaker; (3) state when and where the statement was made; (4) plead with particularity the contents of the false representation; (5) plead with particularity what the person making the misrepresentation obtained thereby; and (6) explain the reason or reasons why the statement is misleading, i.e., why the statement is fraudulent.") (citing ABC Arbitrage Grp. v. Tchuruk , 291 F.3d 336, 350 (5th Cir.2002) ).

Fraud based on an omission must sufficiently allege "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." Matrixx Initiatives, Inc. v. Siracusano , 563 U.S. 27, 38, 131 S.Ct. 1309, 179 L.Ed.2d 398 (2011) (quotation marks omitted). But "[s]ome information is of such dubious significance that insistence on its disclosure may accomplish more harm than good." TCS Indus., Inc. v. Northway, Inc. , 426 U.S. 438, 448, 96 S.Ct. 2126, 48 L.Ed.2d 757 (1976). "[I]f the standard of materiality is unnecessarily low, not only may the corporation and its management be subjected to liability for insignificant omissions or misstatements, but also management's fear of exposing itself to substantial liability may cause it simply to bury the shareholders in an avalanche of trivial information a result that is hardly conducive to informed decisionmaking." Id. at 448–49, 96 S.Ct. 2126. To balance these concerns, "the plaintiff must plead not only why the statement at issue is incomplete but why that incompleteness makes the statement misleading or untrue." R2 Invs. LDC v. Phillips , 401 F.3d 638, 642 (5th Cir.2005). "[I]t bears emphasis that § 10(b) and Rule 10b–5(b) do not create an affirmative duty to disclose any and all material information. Disclosure is required under these provisions only when necessary ‘to make...statements made, in the light of the circumstances under which they were made, not misleading.’ " Matrixx , 563 U.S. at 44, 131 S.Ct. 1309 (quoting 17 C.F.R. § 240.10b–5 ).

While these cases make clear that "[s]ilence, absent a duty to disclose, is not misleading under Rule 10b-5," Basic v. Levinson , 485 U.S. 224, 239 n. 17, 108 S.Ct. 978, 99 L.Ed.2d 194 (1988), "[t]he omission of a known risk, its probability of materialization, and its anticipated magnitude, are usually material to any disclosure discussing the prospective result from a future course of action," Lormand , 565 F.3d at 248. The Fifth Circuit has

long held under Rule 10b-5, a duty to speak the full truth arises when a defendant undertakes a duty to say anything. Although such a defendant is under no duty to disclose every fact or assumption underlying a prediction, he must disclose

material, firm-specific adverse facts that affect the validity or plausibility of that prediction.

Id.

"In cases under the PSLRA, plaintiffs may rely on confidential witnesses provided they are described in the complaint with sufficient particularity to support the probability that a person in the position occupied by the source would possess the information alleged." Owens , 789 F.3d at 542 n. 11 (quotation marks omitted). The court must consider "each [confidential-witness] allegation for its particularization of fraud in compliance with the requirements of Rule 9(b) and section 78u–4(b)(1)." Tchuruk , 291 F.3d at 354.

3. Scienter

Under the PSLRA, plaintiffs must state with particularity facts giving rise to a "strong inference" that the defendant acted with scienter. 15 U.S.C. § 78u–4(b)(2)(A). For "each act or omission alleged" to be false or misleading, plaintiffs must "state with particularity facts giving rise to a strong inference that the defendant acted with the required state of mind." 15 U.S.C. § 78u–4(b)(2)(A). Scienter requires "an intent to deceive, manipulate or defraud or that severe recklessness in which the danger of misleading buyers or sellers is either known to the defendant or is so obvious that the defendant must have been aware of it." Phillips , 401 F.3d at 643 (quotation marks omitted). "Severe recklessness is limited to those highly unreasonable omissions or misrepresentations that involve not merely simple or even inexcusable negligence, but an extreme departure from the standards of ordinary care." Id. (quotation marks omitted).

In Tellabs, Inc. v. Makor Issues & Rights Ltd. ("Tellabs I "), 551 U.S. 308, 127 S.Ct. 2499, 168 L.Ed.2d 179 (2007), the Supreme Court described how to analyze the sufficiency of scienter allegations on a motion to dismiss a federal securities-fraud case under the PSLRA.³ Id. at 322–23, 127 S.Ct. 2499. First, the district court must determine the factually sufficient complaint allegations and take them as true. Id. at 322, 127 S.Ct. 2499. Second, the court considers documents incorporated in the complaint by reference and matters subject to judicial notice. Id. at 322–23, 127 S.Ct. 2499. Third, the court must take into account plausible inferences opposing as well as supporting an inference of scienter. Id. at 323, 127 S.Ct. 2499. The factual allegations must be evaluated collectively, not in isolation, to determine whether a strong inference of scienter has been pleaded. Barrie v. Intervoice – Brite, Inc ., 397 F.3d 249, 260 (5th Cir.2005) ("While [the Fifth Circuit] will view a complaint in toto when considering whether a complaint has adequately pled scienter, each allegation of fraud must individually meet the particularity requirements of the PSLRA." (citation omitted)). "A district court may best make sense of scienter allegations by first looking to the contribution of each individual allegation to a strong inference of scienter, especially in a complicated case...." Owens , 789 F.3d at 537. "As a matter of efficiency, if any single allegation, standing alone, create[s] a strong inference of scienter, the court [does] not need to consider additional allegations of scienter." Id. If analyzing each allegation alone does not support a strong inference of scienter, "the court must follow this initial step with a holistic look at all the scienter allegations," or may assess the allegations holistically, without first engaging in an allegation-by-allegation analysis. Id.

3 The Supreme Court did not decide whether recklessness is sufficient for civil liability under § 10(b) and Rule 10b-5 because the question was not presented in that case. The Court noted that the courts of appeals all permit § 10(b) claims based on reckless behavior but define "recklessness" differently. Tellabs I , 551 U.S. at 319 n. 1, 127 S.Ct. 2499.

The scienter inference need not be "irrefutable, i.e., of the ‘smoking-gun’ genre, or even the most plausible of competing inferences." Tellabs I , 551 U.S. at 324, 127 S.Ct. 2499 (quotation marks omitted). But it "must be more than merely ‘reasonable’ or ‘permissible’—it must be cogent and compelling, thus strong in light of other explanations." Id. "The strength of an inference cannot be decided in a vacuum." Id. at 323, 127 S.Ct. 2499. "To determine whether the plaintiff has alleged facts that give rise to the requisite ‘strong inference’ of scienter, a court must consider plausible, nonculpable explanations for the defendant's conduct, as well as inferences favoring the plaintiff." Id. at 323–24, 127 S.Ct. 2499. "[O]missions and ambiguities count against inferring scienter, for plaintiffs must ‘state with particularity facts giving rise to a strong inference that the defendant acted with the required state of mind.’ " Id. at 326, 127 S.Ct. 2499 (quoting 15 U.S.C. § 78u–4(b)(2) ). "A complaint will survive...only if a reasonable person would deem the inference of scienter cogent and at least as compelling as any opposing inference one could draw from the facts allege." Id. at 324, 127 S.Ct. 2499. "[A] tie favors the plaintiff." Lormand , 565 F.3d at 254.

Courts in this circuit look to the state of mind of the individual corporate official who allegedly made, approved, or issued the statement at issue, or who furnished information or language to include in the statement. It is insufficient to impute to each defendant a kind of collective knowledge of all the corporation's officers and employees acquired in their employment. Flaherty & Crumrine Preferred Income Fund, Inc. v. TXU Corp ., 565 F.3d 200, 208 (5th Cir.2009). The allegations claimed to show scienter on each defendant's part must be analyzed individually to determine whether the complaint sufficiently pleads scienter as to that defendant. Id. ; see also Ind. Elec. Workers' Pension Trust Fund IBEW v. Shaw Group, Inc ., 537 F.3d 527, 532–33 (5th Cir.2008) ("[T]his court has rejected the group pleading approach to scienter....Consequently, it is only necessary for us to address the allegations claimed to adequately show scienter on the part of the named officers to determine whether the complaint sufficiently pleads scienter." (quotation marks omitted)).

The rejection of group pleading requires plaintiffs pleading fraud claims against individuals under § 10(b) and Rule 10b-5 to distinguish among the defendants and allege each one's role, intent, and knowledge. Southland , 365 F.3d at 365 (courts may not "construe allegations contained in the Complaint against the defendants as a group as properly imputable to any particular individual defendant unless the connection between the individual defendant and the allegedly fraudulent statement is specifically pleaded.").

The Fifth Circuit has recently reaffirmed the importance of pleading a strong inference of scienter. In Local 731 I.B. v. Diodes, Inc. , 810 F.3d 951 (5th Cir.2016), the court affirmed the dismissal of a § 10(b) putative class action. The court stressed that "allegations of motive and opportunity standing alone will not suffice, though such circumstantial evidence can enhance the strength of the inference of scienter." Id. at 957 (quotation marks omitted). Without more, "an officer's position with a company does not suffice to create an inference of scienter." Id. at 958 (quotation marks omitted). An officer's position within a company can support the scienter inference only when viewed with other, "special circumstances." Id. at 959. These include (1) a small company in which corporate executives are likely to be familiar with day-to-day operations, (2) transactions "critical to the company's continued vitality," (3) omitted information readily apparent to the speaker, and (4) internally inconsistent statements by the officer. Id .

"Confidential source statements are a permissible basis on which to make an inference of scienter." Integrated Elec. , 497 F.3d at 552. The PSLRA does not require a plaintiff to provide names when the confidential witnesses "are identified through general descriptions in the complaint with sufficient particularity to support the probability that a person in the position occupied by the source as described would possess the information pleaded to support the allegations of false or misleading statements." Tchuruk , 291 F.3d at 353.

In lieu of names, however, a complaint should give details like (1) the person's job description, (2) individual responsibilities, (3) specific employment dates, and (4) where and when the confidential source came to know the information supporting an inference of scienter, such as when a relevant comment was made to the confidential source. See Integrated Elec. , 497 F.3d at 552 ; Shaw , 537 F.3d at 538 ; Southland , 365 F.3d at 382.

Even when a complaint sets out this type of information, the Fifth Circuit has applied Tellabs I to require district courts to "discount allegations from confidential sources" because these "sources afford no basis for drawing plausible competing inferences." Shaw , 537 F.3d at 535 (5th Cir.2008) ; see also Higginbotham v. Baxter Int'l, Inc. , 495 F.3d 753, 757 (7th Cir.2007) ("It is hard to see how information from anonymous sources could be deemed ‘compelling’ or how we could take account of plausible opposing inferences. Perhaps these confidential sources have axes to grind. Perhaps they are lying. Perhaps they don't even exist."). The Fifth Circuit approach requires giving relatively less weight to the allegations from anonymous sources.

4. Loss Causation

Under the PSLRA, "the plaintiff shall have the burden of proving that the act or omission of the defendant...caused the loss for which the plaintiff seeks to recover damages." 15 U.S.C. § 78u–4(b)(4). "Reliance by the plaintiff upon the defendant's deceptive acts is an essential element of the § 10(b) private cause of action." Erica P. John Fund, Inc. v. Halliburton Co. , 563 U.S. 804, 131 S.Ct. 2179, 2184, 180 L.Ed.2d 24 (2011) (quotation marks omitted). The PSLRA "makes clear Congress' intent to permit private securities fraud actions for recovery where, but only where, plaintiffs adequately allege and prove the traditional elements of causation and loss." Dura Pharm., Inc. v. Broudo , 544 U.S. 336, 346, 125 S.Ct. 1627, 161 L.Ed.2d 577 (2005). "To establish proximate causation, the plaintiff must allege that when the ‘relevant truth’ about the fraud began to leak out or otherwise make its way into the marketplace, it caused the price of the stock to depreciate and, thereby, proximately caused the plaintiff's economic harm." Pub. Emps. Ret. Sys. of Miss. v. Amedisys, Inc. , 769 F.3d 313, 320 (5th Cir.2014). The complaint must plausibly allege that the defendant's share price fell "after the truth became known"; alleging "purchase price inflation alone [is not] sufficient." Dura , 544 U.S. at 347, 125 S.Ct. 1627. "Loss causation in fraud-on-the-market cases can be demonstrated circumstantially by (1) identifying a ‘corrective disclosure’ (a release of information that reveals to the market the pertinent truth that was previously concealed or obscured by the company's fraud); (2) showing that the stock price dropped soon after the corrective disclosure; and (3) eliminating other possible explanations for this price drop, so that the factfinder can infer that it is more probable than not that it was the corrective disclosure—as opposed to other possible depressive factors—that caused at least a ‘substantial’ amount of price drop." Amedisys , 769 F.3d at 320–21 (quotation marks omitted).

C. The Section 20 Claims Against Cannon, Karnes, and Khosla

Section 20(a) of the Exchange Act imposes joint and several liability on "[e]very person who, directly or indirectly, controls any person liable under any provision of this chapter or of any rule or regulation thereunder" for securities fraud. 15 U.S.C. § 78t(a). "Control person liability is secondary only and cannot exist in the absence of a primary violation." Southland , 365 F.3d at 383.

"The term control...means the possession, direct or indirect, of the power to direct or cause the direction of the management and policies of a person, whether through the ownership of voting securities, by contract, or otherwise." 17 C.F.R. § 230.405. "Control person liability does not require participation in the fraudulent transaction." Heck v. Triche , 775 F.3d 265, 283 (5th Cir.2014). "But a plaintiff must at least show that the defendant had an ability to control the specific transaction or activity upon which the primary violation is based." Id. (quotation marks omitted). "[The Fifth Circuit] has not yet decided whether a plaintiff must show that the alleged controlling person had ‘effective day-to-day control’ or actually exercised his power over the controlled person." Id. at 283 n. 18.

The plaintiffs argue that their control-person-liability allegations meet the Rule 8 pleading standard, but they cite only cases decided before the 2007 Twombly and 2009 Iqbal decisions. The plaintiffs have not cited a recent case in which a court supported its § 20(a) analysis with a pre-2007 approach to notice-pleading requirements. (Docket Entry No. 97 at p. 61). The court must follow the more recent precedents to the extent they differ from earlier cases.

III. Statements About the Demonstration Unit in Houston, Texas

The plaintiffs allege three types of false or misleading statements about the Houston demonstration unit. They are set out below.

• Statements that the demonstration unit had increased its yield from 17 to 67 gallons per bone-dry ton. (Docket Entry No. 86 at ¶¶ 86, 91, 94, 97) ("We have increased our overall process yield of biomass to renewable fuel from approximately 17 gallons per [bone-dry ton] to approximately 67 gallons per [bone-dry ton]."). These statements were made in KiOR's June 2011 registration prospectus, its 2011 Form 10-K, and its May 2012 and August 2012 Forms 10-Q.

• Statements that the production costs would be less than $1.80 per gallon if the demonstration unit was scaled to commercial size. (Docket Entry No. 86 at ¶¶ 86, 91) ("Our proprietary catalyst systems, reactor design and refining processes have achieved yields of renewable fuel products of approximately 67 gallons per [bone-dry ton] in our demonstration unit that we believe would allow us to produce gasoline and diesel blendstocks today at a per-unit unsubsidized production cost below $1.80 per gallon, if produced in a standard commercial production plant with a feedstock processing capacity of 1,500 [bone-dry tons] per day."). These statements were made in KiOR's June 2011 registration prospectus and its 2011 10-K.

• Sarbanes-Oxley Act certifications affirming, among other things, that Cannon and Karnes had reviewed KiOR's filings and that those filings were free from material misstatements and fairly presented the company's financial condition. (Docket Entry No. 86 at ¶¶ 92, 95, 98).

A. The Safe Harbor

The first question is whether the challenged statements about the Houston demonstration unit are forward looking. "We have increased our overall process yield of biomass...to approximately 67 gallons per [bone-dry ton]" is not a forward-looking statement; rather, it describes what happened in the past. This statement is not protected under the safe-harbor provision. By contrast, the statements about producing gasoline blendstocks at less than $1.80 per gallon are forward looking. The statements use the verb "believe" and the conditional construction that KiOR "would " produce blendstocks for less than $1.80 per gallon "if " KiOR had "a standard commercial production plant with a feedstock processing capacity" of 1,500 bone-dry tons per day. These are predictive "statement[s] of the plans and objectives of management for future operations." See 15 U.S.C. § 78u–5(i)(1)(B).

The next question is whether these statements were identified as forward looking and accompanied by meaningful cautionary language. Both the 2011 registration prospectus and the 2011 Form 10-K identified "statements regarding our future results of operations and financial position, business strategy and plans and our objectives for future operations [as] forward-looking statements. The word[ ] ‘believe’...[is] intended to identify forward-looking statements." (Docket Entry No. 94, Ex. 3 at p. 194, 372). These statements included "the expected production costs and cost-competitiveness of our gasoline and diesel blendstocks." (Id. ).

The prospectus and the Form 10-K set out meaningful cautions, including the following:

We have no experience producing cellulosic gasoline and diesel at the scale needed for the development of our business or in building the facilities necessary for such production, and we will not succeed if we cannot effectively scale our proprietary technology platform and process design.

We must demonstrate our ability to apply our proprietary technology platform and process design at commercial scale to convert biomass into cellulosic gasoline and diesel on an economically viable basis. Such production will require that our proprietary technology platform and process design be scalable from our demonstration unit to commercial production facilities. We have not yet completed construction of or operated a commercial-scale production plant, and our technology may not perform as expected when applied at the scale that we plan or we may encounter operational challenges for which we are unable to devise a workable solution. In particular, our initial-scale commercial production plant under construction in Columbus, Mississippi is a first-of-kind project, and we cannot assure you that it will be completed on the schedule that we intend or at all. If and when completed, our initial-scale commercial production plant may not process biomass at designed levels or produce our cellulosic gasoline and diesel at acceptable yields, and we may be unable to improve its performance. As a result of these risks, we may be unable to achieve commercial-scale production in a timely manner, or at all. If these risks materialize, our business and ability to commercialize

our cellulosic gasoline and diesel would be materially and adversely affected.

The actual cost of constructing, operating and maintaining the facilities necessary to produce our renewable transportation fuels in commercial volumes may be significantly higher than we plan or anticipate.

The production of commercial volumes of our renewable transportation fuels will require the construction of commercial-scale facilities. The construction of these new facilities will require the expenditure of significant amounts of capital, which may exceed our estimates. We may be unable to complete these facilities at the planned costs, on schedule or at all. The construction of new facilities may be subject to construction cost overruns due to labor costs, labor shortages or delays, costs of equipment and materials, weather delays, inflation or other factors, which could be material. In addition, the construction of our facilities may be subject to the receipt of approvals and permits from various regulatory agencies. Those agencies may not approve the projects in a timely manner or may impose restrictions or conditions on a production plant that could potentially prevent construction from proceeding, lengthen its expected completion schedule and/or increase its anticipated cost.

If and when our facilities are constructed, our operating and maintenance costs may be significantly higher than we anticipate. In addition, our facilities may not operate as efficiently as we expect and may experience unplanned downtime, which may be significant. As a result, our initial-scale commercial production plant under construction in Columbus, Mississippi or one or more of the planned standard commercial production facilities may be unable to achieve our expected investment return, which could adversely affect our business and results of operations.

...

We may be unable to realize expected economies of scale, reduce our feedstock costs, increase our overall yields and optimize the composition of our renewable transportation fuels, which could limit our ability to sell our products at competitive prices and materially and adversely affect our business and prospects.

We may be unable to realize expected economies of scale, reduce our feedstock costs, increase our overall yields and optimize the composition of our renewable crude oil in order to produce our renewable fuel products on a cost competitive basis with existing petroleum-based fuel products without government incentives. In particular, we may be unsuccessful in incorporating lower grade woody biomass, such as logging residues, branches and bark, in our process to reduce our feedstock costs or maintain our yields. In addition, our research and development efforts may fail to increase the yield of our [biomass fluid-catalytic-cracking] process such that we may be unable to produce renewable transportation fuels at the costs or in the quantities that we anticipate. Our failure to achieve these efficiencies or improvements over time could limit our ability sell our products at competitive prices and materially and adversely affect our business and prospects.

(Docket Entry No. 94, Ex. 3 at p. 177, 180, 386–87, 390).

These statements in the prospectus and the Form 10-K are meaningful disclaimers. They identified important factors that could make actual results materially different than those predicted in the forward-looking statements, including specific reasons why the commercial-production plant might confront higher production costs than the demonstration unit. The disclaimers warned that the Columbus plant was the first of its kind, that KiOR had "no experience producing cellulosic gasoline and diesel at the scale needed for the development of [its] business," that KiOR had no experience "in building the facilities necessary for such production," and that the business model would "not succeed if [KiOR could not] effectively scale [its] proprietary technology platform and process design." The disclaimers stated that no commercial-scale plant for biofuel production had been built before and that the Columbus production plant's capital costs could exceed expectations. The disclaimers also stated that the plant might "not operate as efficiently as we expect" and might "experience unplanned downtime, which may be significant"; that its core technology would have to be scaled from the demonstration unit; that it might be unable to reach economies of scale necessary to achieve the projected costs; and that research and development might not meaningfully increase the yields at the projected costs.

These prospectus and Form 10-K statements are company-specific, tailored warnings about the Houston demonstration unit that are substantive, meaningful, and relevant to the projection that KiOR could produce blendstocks for less than $1.80 per gallon in a commercial-scale plant. The statements are "company-specific warnings based on a realistic description of the risks applicable to the particular circumstances," see Southland , 365 F.3d at 372, unlike the "boilerplate" warnings rejected in the case law, compare with Lormand , 565 F.3d at 244 ("[The statements are] not guarantees of future performance...and involve known and unknown risks and other factors that could cause actual results to be materially different from any future results expressed or implied by them."), and Plotkin , 407 F.3d at 694 ("These forward-looking statements involve numerous risks, uncertainties and assumptions, and actual results could differ materially from anticipated results.").

The safe-harbor provision bars the plaintiffs' claims based on statements using the Houston demonstration unit to project production costs. The safe harbor does not bar the claims based on KiOR's statements that the demonstration unit had increased its yields from 17 to 67 gallons per bone-dry ton because those statements are not forward looking.

B. Loss Causation

Even if the backward-looking statement that the demonstration unit had achieved increases in KiOR's yields to 67 gallons per bone-dry ton was false, as the plaintiffs allege, there is no accompanying corrective-disclosure allegation. As Cannon observes, the third amended complaint does not allege that the truth about the demonstration unit's past yields was revealed or that a corrective disclosure preceded the stock-price drop. (Docket Entry No. 94 at p. 29 n.27). Because the plaintiffs failed to identify the "release of information that reveal[ed] to the market the pertinent truth that was previously concealed or obscured by the...fraud," the third amended complaint does not state a valid § 10(b) claim based on this alleged misrepresentation. See Amedisys , 769 F.3d at 320–21 ; see also Dura , 544 U.S. at 347, 125 S.Ct. 1627 ("The complaint's failure to claim that [the defendant's] share price fell significantly after the truth became known suggests that the plaintiffs considered the allegation of purchase price inflation alone sufficient," even though "the ‘artificially inflated purchase price’ is not itself a relevant economic loss.").

The statements about the Houston demonstration unit do not give rise to § 10(b) liability. IV. Statements About How the Columbus, Mississippi Plant Operated

A. Representations of Then-Current Facts

The plaintiffs allege the following false or misleading statements about the Columbus plant operations:⁴

4 Cannon and Karnes have not challenged any of these statements as immaterial puffery. The two statements argued to be immaterial are discussed below in Part IV.C.

• On a November 2012 earnings call, Cannon stated: "Our proprietary technology is operating as designed at commercial scale and producing a high quality renewable crude oil, conforming to our design specs....All of this gives me confidence that we will share of the world's first cellulosic gasoline and diesel fuel products on schedule later this month." (Docket Entry No. 86 at ¶ 100).

• On a November 2012 earnings call, Cannon stated: "We've had normal kind of start-up things, the solenoid valve on the cogeneration unit went out and shut us down. Things like that, we've had leaks, which is—on the flue gas—which is very typical of start-up, pumps seals and other things. Just very, very typical." (Id. at ¶ 101).

• In a November 2012 press release, Cannon stated: "The plant's performance to date not only meets our expectations based on our experience at our pilot and demonstration scale facilities, but also gives me confidence that we remain on track to upgrade our oil in order to ship America's first truly sustainable cellulosic gasoline and diesel for American vehicles." (Id. at ¶ 103).

• In a March 2013 press release, Cannon stated: "This plant demonstrates the efficacy of KiOR's proprietary catalytic biomass-to-fuel process with the potential to deliver cellulosic gasoline and diesel to the U.S." (Id. at ¶ 106).

• On a March 2013 earnings call, Cannon stated: "Did we encounter unexpected start-up issues unrelated to our technology [in 2012]? Yes, we did. However, we have overcome these normal start-up issues and we have proven that KiOR's proprietary biomass to fuels technology works at commercial scale at Columbus." (Id. at ¶ 107).

• On a March 2013 earnings call, Cannon stated: "I think probably in the big picture they're all normal startup issues unrelated to our technology....[They're] normal start-up things of getting [the plant] in optimal operation. But thank goodness we have all that behind us now and we've made fuel." (Id. at ¶ 110).

• On a March 2013 earnings call, Cannon stated: "Now, what we're seeing [at the] Columbus [plant] is just operability issues and that can be a pulp. It can be a control valve, other things that are very, very normal when you're starting up a plant. Technology, the chemistry, we couldn't be more happy about. Every time the converter runs it makes a very, very high quality, beautiful oil...." (Id. at ¶ 111).

• In a May 2013 press release, Cannon stated: "[O]ur core technology is continuing to operate as designed, and the plant is producing high quality oil and cellulosic fuel." (Id. at ¶ 114).

• On a May 2013 earnings call, Cannon stated: "Our core technology continues to perform at or above our expectations....If we identify something that needs to be fixed or calibrated,

we do so with the utmost concern for the safety of our employees and the long-term condition of the plant. This means that a typical shutdown for Columbus right now in a start-up phase, meaning going in and making an adjustment that would not normally be considered very significant, can range from a few days to a couple of weeks in duration....Approximately 70 percent of the downtown during the quarter has been associated with addressing typical issues in the vapor-recovery unit or [vapor-recovery unit] at the plant. I call them typical because our [vapor-recovery unit] is at the back end of our conversion process. So any variability and operating conditions upstream, which happen on a frequent basis during a start-up, shows up in the form of plugging in this [vapor-recovery unit]. This is just what you would expect to see in any petrochemical or refining operation like ours.... Understanding this cause-and-effect is important because issues in the [vapor-recovery unit] reflect the impact of typical variable start-up operations and minor problems elsewhere in the plant....The good news is that these issues are not indicative of a major design issue or an issue with our core [biomass fluid-catalytic-cracking] technology....The remaining downtime, approximately 30 percent, was driven by a one-time mechanical issues in utilities and wood processing systems of the plant." (Id. at ¶ 115).

• On a May 2013 earnings call, Cannon stated: "It's certainly not a design, the redesign needed. It's a lot of instrument failures, operating variability, in the front end of the plant, kind of normal start-up issues and they manifest themselves in the vapor-recovery unit. But we've operated it 2 weeks, ran beautifully. So it's not a structural issue." (Id. at ¶ 116).

• On a May 2013 earnings call, Karnes stated: "As important though is the absolute decrease [in start-up costs] itself, the nature of the start-up cost clearly reflects the leveling out and stabilization of the plant as it's handed over from start-up to the operations group." (Id. at ¶ 119).

• At a May 2013 investor presentation, Karnes stated: "And then what we've seen is, the good news is, all the complications that we've incurred have been minor. None of them have implicated any of KiOR's technology." (Id. at ¶ 123).

• In a July 2013 press release, Cannon stated: "The success of these efforts gives us confidence, more than ever, that the performance targets for the Columbus plant are attainable in the months ahead....[and that the plant is] operating stably and producing cellulosic gasoline and diesel at commercial scale." (Id. at ¶ 127).

• On an August 2013 earnings call, Cannon stated: "We decided to terminate [test runs of the plant] due to feed synchronization issues. Nothing about the KiOR technology prevented the runs from going longer....To reiterate, nothing about the KiOR technology prevented this run from going even longer.... As has been the case since we first started the plant, these issues are not related to our core technology. They are simply part of the break-in process....We are now able to operate the plant through issues that would have previously resulted in a shutdown of the plant, and we continue to prove that our proprietary technology works consistent with or better than our expectations." (Id. at ¶ 133).

• On an August 2013 earnings call, Karnes stated: "Again, the issues that the Columbus plant has been working through are not at all related to our technology." (Id. at ¶ 134).

• On a November 2013 earnings call, Cannon stated: "As you know, we kept the [biomass fluid-catalytic-cracking system] down during July, so we can make some repairs and improvements in the Wood Yard....We encountered some issues in the Wood Yard again. Nothing had to do with our [biomass fluid-catalytic-cracking system] or proprietary technology, and brought the [biomass fluid-catalytic-cracking system] down to address these Wood Yard issues on August 24." (Id. at ¶ 139).

• On a November 2013 earnings call, Cannon stated: "[W]e believe that uncertainty and up-and-down operations that marked the start-up period are now largely behind us. We believe that the transition to steady state is increasingly behind us and we are now transitioning to a true operating company." (Id. at ¶ 139).

• On a November 2013 earnings call, Cannon stated: "[Operational issues have] been mostly [in] the Wood Yard. Our last couple of outages were certainly a result of the Wood Yard. We've taken it down, made some improvements there in design and operational improvements and control. So we're...and we're still working there. So I would say in our last quarter and so far, this quarter, of course, so far this quarter, we've done very, very well. But that would be more related to the Wood Yard. The converter has run very [indiscernible]." (Id. at ¶ 140).

• On a November 2013 earnings call, Cannon stated: "The plant we're feeding at 50, 60 percent, maybe even more on a consistent basis now. And what we'll be doing is, of course, getting that up to 100 percent of feed right to the unit. At the same time, we'll be going through our optimization to get the yields up. And it's a process we're still working through. Some...I wouldn't say that the [biomass fluid-catalytic-cracking system] has run very well, so we're still working through to get our Wood Yard up to that level, so we can consistently feed it....[We] will be working over the next few months to accomplish that." (Id. at ¶ 141).

• On a January 2014 earnings call, Cannon described mechanical improvements the company would make to the Columbus plant: "This project is designed to address the three key drivers of Columbus's performance: throughput, yield, and overall process efficiency and reliability.... On throughput, we have identified several improvements that will implement some changes to the [biomass fluid-catalytic-cracking] hydrotreater and wood yard that we believe will eliminate structural design bottlenecks and reliability issues that limit the amount of wood that we can introduce to our [biomass fluid-catalytic-cracking] system....On yield, we have identified additional enhancements that we believe will improve the overall yield of transportation fuels from each ton of biomass from the Columbus plant....On process efficiency and reliability, we have identified another series of improvements that we believe will further optimize our process and increase reliability and on-stream percentage through the plant." (Id. at ¶ 150).

1. The Safe Harbor

None of these statements is forward looking. The statements "referred to then-present factual conditions" about how the technology was operating and whether the problems were merely start-up glitches or more fundamental, structural issues implicating KiOR's core proprietary technology. See Plotkin , 407 F.3d at 699. The statements discussed what had already happened, or what was happening, at the Columbus plant. They did not predict future production, yields, or costs. Statements describing past or present conditions are not forward-looking statements that trigger the PSLRA's safe harbor. Id. (statements about contracts that were already signed were not forward looking); see also Spitzberg , 758 F.3d at 692 ("[W]hether actual production or formation tests have already taken place in the past is undoubtedly backward-looking."); cf. Stone v. Life Partners Holdings, Inc. , 26 F.Supp.3d 575, 597 (W.D.Tex.2014) ("[S]tatements about the present strength of the company and of its product" were not forward looking.).

2. Falsity

The plaintiffs argue that these statements "materially misled investors to believe that the Columbus plant's ‘start-up phase’ was progressing as designed and intended." They also argue that "the statements omitted material information concerning the then-existing atypical, significant design and mechanical flaws affecting operations at the Columbus plant." (Docket Entry No. 86 at ¶¶ 104, 112, 120, 124, 128, 135, 142).

The individual defendants respond that they disclosed technical problems at the Columbus plant but had no legal obligation to characterize those problems as "severe design flaws" rather than "start-up issues." Recognizing that "the defendants disclosed certain ‘start-up’ problems," the plaintiffs contend that the disclosures misled investors by downplaying the problems' severity. (Docket Entry No. 97 at p. 32).

A corporation has "no duty to cast its business in a pejorative, rather than a positive, light." Rosenzweig , 332 F.3d at 869. "[A]s long as public statements are reasonably consistent with reasonably available data, corporate officials need not present an overly gloomy or cautious picture of the company's current performance." Abrams v. Baker Hughes, Inc. , 292 F.3d 424, 433 (5th Cir.2002). Instead, the plaintiffs must plead with particularity the reasons why the defendants created "an impression of a state of affairs that differ[ed] in a material way from [the] one that actually existed." See Shaw , 537 F.3d at 541 (quoting Brody v. Transitional Hosps. Corp. , 280 F.3d 997, 1006 (9th Cir.2002) ).

Cannon and Karnes disclosed during the earnings calls many of the same problems the confidential witnesses identified. The statements made in the earnings calls updated investors about the length of time the Columbus plant operated. (Docket Entry No. 94, Ex. 3 at p. 730; Docket Entry No. 94, Ex. 4 at p. 63, 150). Cannon and Karnes also identified problems that caused plant production shutdowns or delays. Consistent with the allegations in the complaint about the information CW 2 provided, Cannon stated during the November 2012 earnings call that the Columbus plant was running only for "short periods" of no more than three days of continuous production. (Compare Docket Entry No. 94, Ex. 3 at p. 519, with Docket Entry No. 86 at ¶ 42–43). Cannon, like CW 2, described leaks at the plant. (Compare Docket Entry No. 94, Ex. 3 at p. 519, with Docket Entry No. 86 at ¶ 43). During the March 2013 call, Cannon discussed a Columbus-wide "total blackout" that shut down the plant and caused problems for "weeks," including pluggage. This was consistent with what CW 2 revealed about the November 2012 blackouts that shut the plant down. (Compare Docket Entry No. 94, Ex. 3 at p. 582, with Docket Entry No. 86 at ¶ 43).

Similarly, Cannon reported major problems with the vapor-recovery unit, consistent with the allegations in the complaint about the information CW 1 and CW 4 provided. On the May 2013 call, Cannon attributed 70 percent of the plant's downtime in the first quarter of 2013 to plugging in the vapor-recovery unit, similar to CW 1's description of the unit as a "persistent, ongoing problem" and CW 4's description of "mechanical and design problems" at the plant. (Compare Docket Entry No. 94, Ex. 3 at p. 730, with Docket Entry No. 86 at ¶¶ 40–41, 46). Cannon also disclosed that the plant operated 20 percent of the time during the first quarter, again consistent with CW 2's statements about the plant shutting down. (Compare Docket Entry No. 94, Ex. 3 at p. 730, with Docket Entry No. 86 at ¶ 43). Cannon stated that 30 percent of the downtime was due to issues with the wood-processing systems, consistent with CW 2's allegations that the woodchip-feeding system could not "maintain constant operations" and "was unable to properly supply the woodchips," causing shutdowns. (Compare Docket Entry No. 94, Ex. 3 at p. 731, with Docket Entry No. 86 at ¶ 43). This was also consistent with the information from CW 3, CW 4, and CW 5 about the woodchip-feeding-system problems. (Docket Entry No. 86 at ¶¶ 44, 46, 47).

On the May 2013 earnings call, Cannon also reported a failure in the "waste heat boiler," similar to CW 2's statement that the plant "could not handle the wastewater byproduct being produced." (Compare Docket Entry No. 94, Ex. 3 at p. 737, with Docket Entry No. 86 at ¶ 43). CW 3 stated that the plant did not operate in July 2013; so did Cannon. (Compare Docket Entry No. 94, Ex. 4 at p. 150, with Docket Entry No. 86 at ¶ 44). CW 5 reported substantial problems with the woodchip-conveyer systems in September; again, so did Cannon. (Compare Docket Entry No. 94, Ex. 4 at p. 150, with Docket Entry No. 86 at ¶ 47).

To the extent the allegations about the information obtained from the confidential witnesses track the disclosures made in the earnings calls, the plaintiffs have not alleged that the statements misled investors. Rather, the allegations are that Cannon and Karnes provided accurate information about the Columbus plant issues. And to the extent the allegations mirror the reasons KiOR gave for shutting the plant down, (Docket Entry No. 86 at ¶¶ 150, 158–59), the plaintiffs have not plausibly or particularly alleged that the statements were fraudulent when made. Instead, the plaintiffs have at most alleged fraud by hindsight. Courts treat this as insufficient because it is based on " ‘the fact that something turned out badly must mean [the] defendant[s] knew earlier that it would turn out badly.’ " Lormand , 565 F.3d at 254 (quoting Miss. Pub. Emps.' Ret. Sys. v. Boston Scientific Corp. , 523 F.3d 75, 91 (1st Cir.2008) ); Shaw , 537 F.3d at 536 ("At one time the firm bathes itself in a favorable light. Later the firm discloses that things are less rosy. The plaintiff contends that the difference must be attributable to fraud. ‘Must be’ is the critical phrase, for the complaint offers no information other than the differences between the two statements of the firm's condition.") (quoting DiLeo v. Ernst & Young , 901 F.2d 624, 627 (7th Cir.1990) ). The plaintiffs must instead plausibly allege "contemporaneous evidence...that defendants knew earlier what they chose not to disclose until later," id. (quotation marks omitted), because "company officials should not be held responsible for failure to foresee future events," Abrams , 292 F.3d at 433. The plaintiffs' theory is that Cannon and Karnes misrepresented the severity of the problems at the Columbus plant by downplaying them as normal "start-up" issues that did not implicate KiOR's core technology or impede the plant's ability to operate at commercial scale. (Docket Entry No. 97 at p. 8). The theory is that these defendants misled the plaintiffs into believing that KiOR "was a viable business" when it was not. (Id. ). The plaintiffs contend that long before the Columbus plant started operations, the Houston demonstration unit had shown design flaws and low yields, indicating severe obstacles to scaling the technology to commercial size.

CW 6 alleged, for example, that in 2010, he told Khosla that KiOR's technology "was not yet fully developed" and that the company should not go public. (Docket Entry No. 86 at ¶ 49). When KiOR released the 2011 prospectus, CW 6 received calls from "colleagues" questioning "the accuracy of the information relating to the company's technology; specifically, the prospectus appeared to contain inaccurate information regarding the [demonstration unit's] yield...." (Id. at ¶ 50). CW 6 "was aware of data that indicate[d] yields of less than 40 gallons per ton." (Id. ). "CW 6 was aware of this data because of CW 6's involvement with the technology and, specifically, CW 6's involvement with the demonstration plant in Houston, Texas." (Id. ). CW 6 "monitored the research and development at the demonstration plant, evaluated data generated by the demonstration unit, and worked full-time at the demonstration plant." (Id. ). After the prospectus was released, CW 6 attended an investor conference and heard KiOR officials discussing future target yields at the Columbus plant. The KiOR officials set target yields of 72 to 90 gallons, even though, according to CW 6, "the demonstration unit was showing maximum yields of 42 gallons per ton." (Id. at ¶ 51).

In September 2011, CW 6 complained to KiOR's COO, William Coates, that the "technology was not capable of producing the yields at the costs previously represented." (Id. at ¶ 52). Coates relayed this information to Cannon, but, according to CW 6, Coates "was fired and/or resigned" immediately after the meeting. (Id. ).

Paul O'Connor also stated that in late 2011, after the prospectus release but before the Columbus plant became operational, he received "additional data" showing that "yields were lower than reported in February (and then projected in the [registration prospectus] ) and that hardly any progress had been made since the end of 2009." (Id. at ¶ 64). O'Connor was "shocked." (Id. ). He reported this information to Cannon, who permitted him to conduct a "technology review." O'Connor concluded that yields "had not improved significantly over the last two years" and that while it was "possible to reach the target of 67 gallons [per bone-dry ton]," doing so would require "a drastically different approach[.]" (Id. ). O'Connor also alleged that in an April or May 2012 meeting, KiOR's research and development director "admitted that [KiOR] should not expect to reach" a 67-gallon yield. (Id. at ¶ 65). O'Connor estimated that the yield would be closer to 30 to 40 gallons. (Id. ). CW 9 is also alleged to state that even though problems were "apparent" at the Houston demonstration unit, "[KiOR] never did anything to change it," but rather "ignore[d] [it] and hope[d] it wouldn't happen in Columbus." (Id. at ¶ 59).

The plaintiffs argue that these warning signs became a reality once the Columbus plant began operations. The plaintiffs emphasize the complaint allegations that the plant faced problems more severe than Cannon or Karnes disclosed. The plaintiffs rely, for example, on CW 1's statement that the vapor-recovery unit had "significant design flaws" and needed to be "redesigned." Cannon disclosed problems relating to the vapor-recovery unit during the May 2013 earnings call, but he assured investors that the problems were "typical" and happened "on a frequent basis during a start-up." (Compare Docket Entry No. 86 at ¶ 41, with Docket Entry No. 94, Ex. 3 at p. 730). CW 2 stated that the plant had "operational problems" and "design issues" and that "the whole damn thing was a design issue. It was hastily put together and they were trying to make it run." CW 2 identified undersized equipment that could not handle wastewater byproduct, tar buildup in the fractionating system, leaks in the wood-chip-feeding system, and blackouts as examples of significant, long-term design issues. Although Cannon disclosed many of these issues, he described them as "normal" and part of the start-up process. (See, e.g. , Docket Entry No. 86 at ¶ 110–11). By contrast, CW 2 stated that these design issues "were apparent to everyone" "within just a few days of operations." (Id. at ¶ 43). CW 4 described similar "design issues" with the wood-chip-feeding system, undersized equipment, and the vapor-recovery unit, among others. (Id. at ¶ 46).

The third amended complaint alleges that despite the information and reports showing severe problems at the plant and that KiOR's technology could not operate on a commercial scale, Cannon told investors in November 2012 that the "technology is operating as designed at commercial scale" and that the start-up issues were "normal." (Id. at ¶¶ 100–01). In March 2013, Cannon told investors that the plant "demonstrates the efficacy of KiOR's [technology]" and that "we have overcome these normal start-up issues and we have proven that KiOR's [technology] works at commercial scale." (Id. at ¶¶ 106–07). The "normal start-up issues," Cannon assured investors, were "behind us now." (Id. at ¶¶ 110–11). And even though Cannon told investors about week-long plant shutdowns in May 2013, he insisted that these were start-up issues "not indicative of a major design issue or an issue with our core technology," but "minor" and "one-time mechanical issues." (Id. at ¶¶ 115–16). Karnes also told investors that start-up costs were decreasing and reflected a "leveling out and stabilization of the plant." (Id. at ¶ 123).

Throughout 2013, the complaint alleges, information showing increased barriers to achieving commercial-scale production became evident. CW 6 stated that in mid-2013, a staff engineer, Charlie Zhang, told him that the plant was producing, at most, 22 gallons per bone-dry ton and that costs per gallon were "exceedingly greater than the costs being reported at that point in time." (Id. at ¶ 53). CW 6 wrote a letter in August or September 2013 to Cannon and Christopher Artzer, KiOR's vice-president, stating these concerns. CW 6 then met with Cannon and Artzer in person and told them that he was about to take the concerns to the board of directors. Cannon responded by asking CW 6 to sign a separation agreement. (Id. at ¶ 54).

CW 7 participated in a 52-day production run starting in May 2013. This run made it "evident" that the plant could not meet the yield expectations because of "ongoing operational issues," including "conversion issues arising from the chemistry, catalyst, and wood ratio." (Id. at ¶ 55). CW 7 reported that he was "uncomfortable" with the projections because there was not a "snowball's chance" they would be achieved. (Id. ).

O'Connor expressed concerns about the technology directly to Cannon in May 2013, but he was told he was being "too negative." (Id. at ¶ 66). By November 2013, O'Connor stated that he had become "convinced" that yields were "much lower than projected," even though his first impression was that the plant was experiencing "start-up issues." (Id. ).

Despite these reports of increasingly severe production problems, Cannon and Karnes continued to tell investors that KiOR was "operating stably and producing cellulosic gasoline and diesel at commercial scale" and that the issues were "not related to [KiOR's] core technology." (Id. at ¶¶ 127, 133, 134). In November 2013, Cannon told investors that the "up-and-down operations that marked the start-up period are now largely behind us." (Id. at ¶ 139). But less than two months later, Cannon announced that the Columbus plant would be shut down for the first quarter of 2014. Four months later, the plant was shut down indefinitely. See Plotkin , 407 F.3d at 698 ("[A]llegations of later-emerging facts can, in some circumstances, provide warrant for inferences about an earlier situation.").

The plaintiffs have plausibly alleged, with particularity, that the statements about the Columbus plant's operations materially misled investors to believe that KiOR was a viable business when it was not. From November 2012 to November 2013, Cannon and Karnes told investors that KiOR's core technology was fundamentally sound, that the Columbus plant operated successfully on a commercial scale, and that any problems were normal, start-up glitches rather than fundamental, structural defects. The third amended complaint plausibly alleges that those representations painted a misleading picture of conditions at the Columbus plant. CW 6, CW 9, and Paul O'Connor allegedly stated that their personal experience with, or information about, the Houston demonstration unit showed that the technology could not be replicated at a commercial-scale production plant.

These and other confidential witnesses, all former KiOR employees, detailed issues at the Columbus plant throughout this period that were more severe than Cannon and Karnes represented. Start-up costs and net-operating losses consistently increased. (Docket Entry No. 86 at ¶¶ 181–82). CW 6, CW 7, and Paul O'Connor allegedly stated that yields were lower, and costs higher, than represented, conflicting with Cannon's and Karnes's representations that KiOR had proved that its technology worked on a commercial scale. Cf. Lormand , 565 F.3d at 249 ("The defendants continually skewed the mix of information by omitting the known severe risks associated with these business actions even as they recognized signs that those risks had already materialized."). And even though Cannon disclosed some of these issues in January 2014, he also assured investors on the earnings call that the company would make improvements to the plant that would have it back up and running. The plant was shut down indefinitely two months later. See Plotkin , 407 F.3d at 698.

Cannon and Karnes respond that the plaintiffs' factual allegations primarily rest on semantics. Cannon and Karnes argue that they had no obligation to characterize the operational issues at the Columbus plant as "severe design flaws" rather than as "start-up issues." (Docket Entry No. 94 at p. 33). But by downplaying the extent and duration of the technological problems as normal start-up issues, while simultaneously assuring investors that the core technology was sound for commercial-scale production, these defendants represented that the Columbus plant was a viable operation that had minor and temporary setbacks, overcame them, and was now ready for the long haul. Even when the plant shut down for the first quarter of 2014, Cannon identified improvements for commercial-scale production to resume operations. The third amended complaint plausibly alleges the misleading nature of these representations with facts that are sufficiently particularized to explain why and on what basis they were misleading.

Cannon and Karnes try to distinguish their statements about "core technology" from the statements highlighted in the third amended complaint, which they argue concern only the Columbus plant "as a whole." At oral argument, they also tried to distinguish their statements about achieving "commercial-scale" production from broader representations that the plant was economically viable, arguing that commercial-scale production referred only to the volume of the gasoline produced rather than to the costs incurred in production. (Docket Entry No. 107 at p. 28). These distinctions cannot be the basis for dismissal at this stage on this record.

The Fifth Circuit's recent decision in Spitzberg rests on similar facts and is instructive. The plaintiffs, investors in a Houston-based energy company, alleged that the company's statement that it had "estimated recoverable reserves of 1 to 4 billion barrels" of oil was misleading. Spitzberg , 758 F.3d at 680. The plaintiffs argued that the statement "communicated to investors that certain geological testing had been completed based on the definition of ‘reserves' used by the oil industry and by SEC regulations." Id. The defendants countered that they had not "explicitly represent[ed] that any drilling had yet occurred," that their other references to drilling were ambiguous, and that "no investor would understand or did understand the term ‘reserves' to mean ‘reserves' as defined by the [oil industry]." Id. at 681.

The Fifth Circuit reasoned that although the defendants "never explicitly represented that they were using the [oil-industry] definition of ‘reserves,’ " the plaintiffs had plausibly alleged that investors understood the term in that context. Id. at 689. The court recognized that "whether [the defendants] will ultimately prevail on their argument that no investor would understand or did understand the term ‘reserves' to mean ‘reserves' as defined by the [oil industry] is another matter." Id. (quotation marks omitted). "That question can only be answered based on the evidence produced during a later stage of th [e] litigation." Id.

Here, the plaintiffs' factual allegations support the inference that an investor would reasonably understand "core technology" to refer to the processes used in the Columbus production plant as a whole, not merely to a specific subpart of the plant's operations, such as the chemical process for "cracking" the molecular structure of the woodchips or the use of a particular piece of equipment. The plaintiffs' factual allegations similarly support the inference that an investor would reasonably understand "commercial-scale" production to refer to the plant's commercial viability, not to the yield volume divorced from the costs of production. As in Spitzberg , whether these terms were reasonably susceptible to different interpretations is properly resolved at later stages of the litigation, not at this stage. See id.

Cannon and Karnes also challenge the confidential witnesses' qualifications to offer opinions about design flaws in KiOR's technology. They argue, for example, that CW 1 and CW 2 were "several levels removed from any operational authority" and could not distinguish between a design flaw and a start-up issue at the Columbus plant, which was a "first-of-a-kind plant." (Docket Entry No. 94 at p. 34 & n.36). But CW 1 and CW 2 worked at the Columbus plant from June 2012 to February 2013. The third amended complaint relies on their firsthand observations from that work. Both CW 1 and CW 2 also had engineering and oil-refinery experience, and CW 2 attended daily operational meetings with the plant manager to discuss mechanical problems that had arisen. This experience is not insufficient, and the witnesses are not unreliable, simply because the plant was the "first of its kind." Taken to its logical conclusion, this argument would mean that no one could offer firsthand observations about an innovative plant, including how it functioned or the nature of its operational problems.

Cannon and Karnes also argue that the allegations of the confidential witnesses' statements and reports are not sufficiently particular. These defendants also criticize the failure to directly quote the confidential witnesses and instead vaguely and conclusorily summarizing the information allegedly obtained from them. A complaint relying in part on confidential witnesses must identify the witnesses' job descriptions, individual responsibilities, specific employment dates, experience, and how they knew the information alleged. See Dawes v. Imperial Sugar Co. , 975 F.Supp.2d 666, 692 (S.D.Tex.2013) (citing Integrated Elec. , 497 F.3d at 552 ; Shaw , 537 F.3d at 538 ; Southland , 365 F.3d at 382 ). The third amended complaint provides this information.⁵ (See Docket Entry No. 97 at p. 32–33).

5 The challenge to the information allegedly provided by CW 3 and CW 5 presents the closest question. (Docket Entry No. 94 at p. 34 n.37). CW 3 worked at the Columbus plant over the summer of 2013 as a millwright. CW 5 worked at the plant for several weeks on the maintenance crew. The court need not decide now whether these allegations are sufficiently particular, or whether these confidential witness are sufficiently qualified, to meet the PSLRA standard because other confidential witnesses and Paul O'Connor meet that standard.

" Rule 9(b) supplements but does not supplant Rule 8(a)'s notice pleading." Kanneganti , 565 F.3d at 185–86. It "does not reflect a subscription to fact pleading and requires only simple, concise, and direct allegations of the circumstances constituting fraud, which after Twombly must make relief plausible, not merely conceivable, when taken as true." Id. at 186 (footnote omitted) (quotation marks omitted). The confidential witnesses describe their firsthand observations of particular equipment that they worked with or observed. They describe deficient equipment, problems in operating the equipment, the severity of these deficiencies and problems, and the consequences for plant operations. The third amended complaint did not need also to specify whether or how the confidential witnesses diagnosed the engineering-or chemistry-based plant problems they reported. The pleading stage did not require the plaintiffs to allege or attach evidentiary support for the confidential witnesses' statements, such as diagrams or mathematical models, demonstrating or explaining how and why the reported equipment failures or problems occurred. As the Fifth Circuit has explained:

The PSLRA was enacted, in part, to compensate for the perceived inability of Rule 9(b) to prevent abusive, frivolous strike suits. It was not enacted to raise the pleading burdens under Rule 9(b) and section 78u–4(b)(1) to such a level that facially valid claims, which are not brought for nuisance value or as leverage to obtain a favorable or inflated settlement, must be routinely dismissed on Rule 9(b) and 12(b)(6) motions....[T]he plaintiffs need not allege all facts that may be related to their claims, since such a requirement is impossible at the pleading stage because, in nearly every securities fraud case, only the defendants know all the facts related to the alleged fraud. In this sense, the PSLRA may have changed federal securities law; it did not eliminate it.

Tchuruk , 291 F.3d 336, 354 (alterations omitted) (footnotes omitted) (quotation marks omitted).

Cannon and Karnes finally argue that any statements about KiOR's "expectations" cannot have been false or misleading because the third amended complaint does not allege what the expectations were in the first place. Case law supports this argument. See, e.g. , Browning v. Amyris, Inc. , No. 13–cv–2209, 2014 WL 1285175, at *11 (N.D.Cal. Mar. 24, 2014) (statement that "[o]ur yeast, our processes and our equipment performed as expected" was not false or misleading because the plaintiffs did not allege "what expectations [the defendant] had concerning its yeast, processes, or equipment during the start-up that did not perform as [the defendant] expected"). These statements are not actionable. But the plaintiffs have plausibly and sufficiently alleged that the remaining statements are false or misleading, precluding dismissal on this basis.⁶

6 The court need not decide whether it is proper to consider the bankruptcy-court testimony from KiOR's vice-president, Christopher Artzer, that the plaintiffs introduced for the first time in their omnibus response to the motions to dismiss. (Docket Entry No. 97 at p. 35–36). As discussed, the complaint allegations are sufficient to allege falsity.

3. Scienter

For "each act or omission alleged" to be false or misleading, the plaintiffs must "state with particularity facts giving rise to a strong inference that the defendant acted with the required state of mind." 15 U.S.C. § 78u–4(b)(2)(A). The plaintiffs allege that on at least six different occasions, O'Connor and CW 6 alerted senior management, including Cannon and Karnes, that KiOR was releasing false or misleading information to the public. The occasions are described in the third amended complaint, as follows:

• At or around the time of the IPO, CW 6 complained to KiOR's executive management, including Cannon and Karnes, in writing and in person about the inaccuracy of the registration-statement sections on the company's yield capabilities, (Docket Entry No. 86 at ¶ 50);

• In September 2011, after resigning from the executive-management team over concerns about misrepresentations, CW 6 met with KiOR's COO, William Coates, to explain that KiOR's technology was not capable of producing the yields represented to the public. Coates would later take this information to Cannon, after which he was "fired and/or resigned immediately." CW 6 continued to complain to KiOR's executive management, including Cannon and Karnes, about the inaccuracy of the yield representations, (id. at ¶ 52);

• In March 2012, after an internal audit, O'Connor gave Samir Kaul and defendant Khosla his final report reflecting accurate yield capabilities at the Columbus plant; in turn, Kaul relayed the information to Cannon who, in reply, advised Kaul that O'Connor was being too pessimistic, (id. at ¶ 72);

• In May 2013, O'Connor sent Cannon and Samir Kaul a letter requesting certain information about the plant's operations and expressing concern that KiOR needed to disclose the plant problems to the public. Cannon responded that O'Connor was "too negative" and that "[KiOR] [had] made tremendous progress in the last 18 months in [research and development]," (id. at ¶ 66);

• In August or September 2013, CW 6 called for a meeting with Christopher Artzer, KiOR's vice-president, and Cannon. In advance of the meeting,

CW 6 sent a letter setting out concerns about misinformation being disseminated. At the meeting, CW 6 shared the information and reported that he would be taking his concerns to KiOR's board of directors. Cannon responded by asking CW 6 to sign a separation agreement, (id. at ¶ 54); and

• In September 2013, O'Connor again questioned Kaul about the Columbus plant production. Kaul responded that he could not reveal this "inside" information because O'Connor was no longer a director, (id. at ¶ 74).

(Docket Entry No. 86 at ¶ 213).

The plaintiffs also rely on other indicia of scienter, including: Cannon's and Karnes's incentive to keep the stock price high and continue collecting salaries; the "core-business" exception, which allows courts to infer what an executive of a small, one-dimensional company "must have known"; insider-trading allegations; Karnes's resignation; and the SEC investigation.

The scienter allegations as to Karnes and Cannon are analyzed in turn below.

a. Karnes

The third amended complaint plausibly alleges that Karnes made the following false or misleading statements:

• On a May 2013 earnings call, Karnes stated: "As important though is the absolute decrease [in start-up costs] itself, the nature of the start-up cost clearly reflects the leveling out and stabilization of the plant as it's handed over from start-up to the operations group." (Id. at ¶ 119).

• At a May 2013 investor presentation, Karnes stated: "And then what we've seen is, the good news is, all the complications that we've incurred have been minor. None of them have implicated any of KiOR's technology." (Id. at ¶ 123).

• On an August 2013 earnings call, Karnes stated: "Again, the issues that the Columbus plant has been working through are not at all related to our technology." (Id. at ¶ 134).

The court will first consider the scienter allegations separately. It will then consider them holistically. See Owens , 789 F.3d at 537. The court finds that considered either individually or together, the allegations do not give rise to a strong inference of scienter as to Karnes.

i. Whether the Confidential-Witness Allegations Give Rise to a Strong Inference of Scienter

The allegations about what the confidential witnesses reported to Karnes are limited to statements from CW 6 claiming that, on two separate occasions, CW 6 told KiOR's executive management, including Cannon and Karnes, that statements they made about the Houston demonstration unit's yields were false. The remaining scienter allegations concern information relayed only to Cannon. These allegations cannot support a strong inference of scienter as to Karnes, in part because the alleged conversations with Karnes antedate his challenged statements by more than a year. See Shaw , 537 F.3d at 535 (confidential-witness allegations alone "afford no basis for drawing the plausible competing inferences required by Tellabs .").

Even if these two statements could support an inference that Karnes recklessly disregarded evidence tending to show significant obstacles to the Columbus plant scaling the Houston demonstration unit's technology to commercial size, that inference is neither cogent nor compelling. See Tellabs I , 551 U.S. at 325, 127 S.Ct. 2499. When Karnes made the challenged statements, the Columbus plant had been operating for several months. The third amended complaint does not allege that Karnes was involved in overseeing the Columbus plant operations. He was the company's chief financial officer. The plaintiffs do not allege that he was responsible for, or involved in, developing or monitoring the technology or its implementation in Houston or Columbus. Yet the plaintiffs seek to draw the inference from the allegations of a single confidential witness that because Karnes was aware of data showing lower-than-reported yields at the Houston demonstration unit, he acted severely recklessly in disregarding that data when he told investors about the state of the Columbus plant several months after it began operating.

There is a plausible, nonculpable explanation for Karnes's statements. It is that he reasonably relied on data showing that the plant yields and operating times were improving. In the second quarter of 2013, for example, the plant doubled its operating-time percentage from 20 to 43 percent and shipped 75,000 gallons of oil. (Docket Entry No. 94, Ex. 4 at p. 63–64). This competing and nonculpable explanation is far stronger than the scienter inference the plaintiffs seek to draw. See Tellabs I , 551 U.S. at 324, 127 S.Ct. 2499 ("[T]he inference of scienter must be more than merely ‘reasonable’ or ‘permissible’—it must be cogent and compelling, thus strong in light of other explanations . A complaint will survive, we hold, only if a reasonable person would deem the inference of scienter cogent and at least as compelling as any opposing inference one could draw from the facts alleged." (emphasis added)).

The plaintiffs purport to incorporate by reference allegations in a related, state-court action filed by the State of Mississippi against Cannon, Khosla, and others. This action seeks to recover money Mississippi loaned KiOR in exchange for investing in a second Mississippi plant similar to the Columbus plant. (Docket Entry No. 86 at ¶ 21). In the state-court complaint, William Coates, KiOR's former COO, described a September 2011 executive-team meeting in which he accused Cannon and Karnes of "cooking the books" based on their representations about costs and yields at the Houston demonstration unit. (Id. at ¶ 52 n.9). The parties dispute whether a plaintiff can incorporate by reference pleading allegations from a case involving different parties. (Compare Docket Entry No. 94 at p. 46 n.48, with Docket Entry No. 97 at p. 35). The court need not resolve this question, because even if Coates's statement to Cannon and Karnes that they were "cooking the books" is considered, it does not give rise to, or support, a strong inference that Karnes acted with the required scienter. The "cooking-the-books" allegation was about the Houston demonstration unit in 2011, but the statements challenged here are about the Columbus, Mississippi plant in May and August 2013. The allegation is also too vague and imprecise to support the type of "compelling" and "cogent" inference the case law requires. See Tellabs I , 551 U.S. at 324, 127 S.Ct. 2499.

The allegations about the information the confidential witnesses provided do not give rise to a strong inference of scienter as to Karnes.

ii. Whether Increasing Debt, Losses, and Start-Up Costs Give Rise to a Strong Inference of Scienter

The plaintiffs allege that in telling investors about the strength of KiOR's operations and technology, Karnes ignored, with severely reckless disregard, increases in KiOR's accumulated debt, net-operating losses, and start-up costs. (Docket Entry No. 86 at ¶¶ 180–83). The plaintiffs contend that the magnitude of these changes gives rise to a strong inference that Karnes made the May and August 2013 statements with scienter. The record shows that Karnes disclosed the "red flags" the plaintiffs identify. He repeatedly informed investors about the status of KiOR's debt, losses, and start-up costs, (see, e.g. , Docket Entry No. 94, Ex. 3 at p. 732–34; Docket Entry No. 94, Ex. 4 at p. 66), and the plaintiffs do not allege that Karnes misrepresented the amounts of start-up costs. When "the red flags [are] disclosed to the public," the disclosure "negates the inference that defendants acted with scienter." Owens , 789 F.3d at 540. "[The] documents referenced in the [third amended complaint] and attached to defendants' motion to dismiss confirm that the alleged red flags...were disclosed promptly...." See id. at 541. The representations alleged to be misleading are those describing the strength and operability of the technology at the Columbus plant, not KiOR's financial information.

The increasing debt, losses, and start-up costs do not give rise to, or support, a strong inference of scienter as to Karnes.

iii. Whether Karnes's Resignation Gives Rise to a Strong Inference of Scienter

The plaintiffs allege that Karnes's "unexpected and unexplained" December 2013 resignation gives rise to a strong inference of scienter. (Docket Entry No. 86 at ¶ 184). Executive resignations are generally "unavailing as proof of the commission of fraud." Southland , 365 F.3d at 383 ; see also Abrams , 292 F.3d at 434 (executive resignations did not have "any scienter implications"); In re Arth r oCare Corp. Sec. Litig. , 726 F.Supp.2d 696, 724–25 (W.D.Tex.2010) ("Multiple Fifth Circuit decisions suggest resignations have little implication on the scienter analysis."). "The resignation of officials is, in and of itself, unavailing as proof of the commission of fraud when no specific evidence indicates the resigning officials or their replacements knew of any [misrepresentations] or that such [misrepresentations] were the reason for their resignations." In re Arth r oCare , 726 F.Supp.2d at 725. This allegation is insufficient to plead scienter as to Karnes.

iv. Whether the "Core-Business" Exception Gives Rise to a Strong Inference of Scienter

The plaintiffs allege that because KiOR's business centered around producing cellulosic gasoline, Karnes's position as CFO made it likely that he was aware of important issues affecting the production. (Docket Entry No. 86 at ¶ 180; Docket Entry No. 97 at p. 50–52). The plaintiffs cite Nathenson v. Zonagen , 267 F.3d 400 (5th Cir.2001), in which the court held that the individual defendants' positions within the defendant pharmaceutical company enhanced the scienter allegations.

In Nathenson , the court also "recognize[d] that normally an officer's position with a company does not suffice to create an inference of scienter." Id. at 424 ; see also Shaw , 537 F.3d at 535 ("[Fifth Circuit] caselaw makes clear that pleading[s] of scienter may not rest on the inference that defendants must have been aware of the misstatement based on their positions with the company." (quotation marks omitted)). The court noted "a number of special circumstances... taken together, [that] suffice[d] to support a different result in the present case." Nathenson , 267 F.3d at 425. The defendant company was small and had only three-dozen full-time employees; it was essentially a one-product company; and the alleged misrepresentations were about the patent protection for that single product, the company's most crucial issue.

The Fifth Circuit and other courts have been reluctant to apply the limited exception recognized in Nathenson . See Rosenzweig , 332 F.3d at 867–68 (rejecting the plaintiffs' argument that "the failure of Azurix's core business—water-privatization projects—supports the inference that defendants knew, or recklessly disregarded, Azurix's prospects for success" and holding that the plaintiffs must "identify exactly who supplied the information or when they knew the information"); Abrams , 292 F.3d at 432 ("A pleading of scienter may not rest on the inference that defendants must have been aware of the misstatement based on their positions within the company.").⁷ Instead, only in the "rare" case will a strong inference of scienter be drawn from an officer's position in a company, and only when this factor combines with other, "special circumstances." Diodes , 810 F.3d at 959. These circumstances include: (1) a small company where it is more likely corporate executives are familiar with day-to-day operations; (2) transactions "critical to the company's continued vitality"; (3) omitted information readily apparent to the speaker; and (4) statements by the corporate officer that are internally inconsistent. Id .

7 See also Collmer v. U.S. Liquids, Inc ., 268 F.Supp.2d 718, 754 (S.D.Tex.2003) ("Plaintiffs have relied on the judicially created presumption that facts critical to a business's core operations or an important transaction generally are so apparent that their knowledge may be attributed to the company and its key officers....[T]he purpose of the PSLRA's particularized pleading requirements leads this Court to find that such an imputation, without some additional facts such as exposure to content-identified internal corporate documents (and who drafted them, who received them or how plaintiffs learned of them) or specific conversations or attendance at specified management or board meetings dealing with such problems, is inadequate to plead scienter." (citations omitted) (quotation marks omitted)).

These circumstances are not present here. KiOR has 183 employees, many more than the companies falling under the Nathenson exception. See Dorsey v. Portfolio Equities, Inc. , 540 F.3d 333, 342 (5th Cir.2008) (no employees); Nathenson , 267 F.3d at 425 (32 to 35 employees) ; (Docket Entry No. 94, Ex. 5 at p. 99). The plaintiffs do not allege that Karnes's statements about KiOR's financial status on the earnings calls were internally inconsistent or misleading. Although the commercial success of KiOR's biofuel-production process was critical to its survival, Karnes was the CFO, not the CEO. His corporate position provides a weaker basis for inferring scienter as a result. The plaintiffs do not allege that Karnes supervised, monitored, or evaluated plant operations or the progress of the KiOR technology. And the third amended complaint does not allege facts permitting a plausible inference that the information Karnes allegedly omitted was "readily apparent" to him.

The plaintiffs have not pleaded a basis to infer that Karnes's culpable knowledge of the plant's day-to-day operations and the production issues can be presumed under the limited circumstances outlined in Nathenson .

v. Whether the SEC Investigation Gives Rise to a Strong Inference of Scienter

The plaintiffs allege that the facts that the SEC is investigating KiOR and has issued subpoenas gives rise to a strong inference of scienter as to Karnes. (Docket Entry No. 86 at ¶ 216). At least two circuit courts have held that an SEC investigation is not enough to support a strong inference of scienter. See Konkol v. Diebold, Inc. , 590 F.3d 390, 402 (6th Cir.2009), abrogated on other grounds by Frank v. Dana Corp. , 646 F.3d 954 (6th Cir.2011) ; In re Ceridian Corp. Sec. Litig. , 542 F.3d 240, 248–49 (8th Cir.2008). "Although a government investigation is not altogether irrelevant to the scienter analysis, a decision by government agencies to investigate a company is not sufficient to meet the heightened Tellabs standard on its own." Konkol , 590 F.3d at 402. "Government investigations can result from any number of causes, and the investors have not pointed to any facts suggesting that the SEC investigation was the result of knowing or reckless behavior by the Defendants." See id. "[T]he opposing inferences—[for example,] that the SEC investigation uncovered no evidence of fraud...—are more compelling in the absence of particular facts giving rise to a strong inference of fraud." In re Ceridian Corp. , 542 F.3d at 248–49. The plaintiffs have not alleged the outcome of the SEC investigation or other facts that give rise to a strong inference of scienter as to Karnes.⁸

8 The plaintiffs also allege that the $185 million impairment charge KiOR disclosed in its 2013 Form 10-K gives rise to a strong inference of scienter as to Karnes. (Docket Entry No. 86 at ¶ 160). "Assuming the amount of the impairment to be true, such allegation sounds in mismanagement or negligence, both of which are insufficient to give rise to a strong inference of scienter." In re Capstead Mortg. Corp. Sec. Litig. , 258 F.Supp.2d 533, 545–55 (N.D.Tex.2003).

vi. Whether Motive-and-Opportunity Allegations Enhance the Strength of the Scienter Inference

"[A]ppropriate allegations of motive and opportunity may meaningfully enhance the strength of the inference of scienter, but... allegations of motive and opportunity, without more, will not fulfill the pleading requirements of the PSLRA." Owens , 789 F.3d at 539 (alterations in original) (quotation marks omitted). The Fifth Circuit "requires more than allegations of motive and opportunity to withstand dismissal." Southland , 365 F.3d at 368 (quotation marks omitted). "To demonstrate motive, plaintiffs must show concrete benefits that could be realized by one or more of the false statements and wrongful nondisclosures alleged. Merely alleging facts that lead to a strained and tenuous inference of motive is insufficient to satisfy the pleading requirement." Shaw , 537 F.3d at 543 (quotation marks omitted).

The plaintiffs claim that three types of motive-and-opportunity allegations enhance support for a strong inference of scienter: insider trading, the desire to raise capital needed to sustain operations, and the incentive to earn executive compensation. Each is separately considered.

A. Insider Trading

The plaintiffs allege that Karnes engaged in "unusual and atypical insider transactions while [KiOR's] stock was artificially inflated during the class period." (Docket Entry No. 86 at ¶ 188). They argue that these stock sales enhance support for a scienter inference for Karnes's May and August 2013 statements about the Columbus plant operations.

"With respect to the requirement that particular facts be pled which give rise to a strong inference of scienter, allegations of insider trading are essentially a form of motive and opportunity allegations." Southland , 365 F.3d at 368. "Because corporate executives are often paid in stock and stock options, they will naturally ‘trade those securities in the normal course of events,’ and courts ‘will not infer fraudulent intent from the mere fact that some officers sold stock.’ " Shaw , 537 F.3d at 543 (quoting In re Burlington Coat Factory Sec. Litig. , 114 F.3d 1410, 1425 (3d Cir.1997) (Alito, J.)). Insider-trading allegations may enhance the strength of the scienter inference "only" when the trades occur "in suspicious amounts or at suspicious times." Southland , 365 F.3d at 368 (quotation marks omitted). "Suspicion may be generated if the sales are out of line with prior trading practices or made at times calculated to maximize personal profit." Shaw , 537 F.3d at 543 (alteration omitted) (quotation marks omitted); see also Southland , 365 F.3d at 368 ("[I]nsider trading is suspicious only when dramatically out of line with prior trading practices at times calculated to maximize the personal benefit from undisclosed insider information." (quoting In Re Silicon Graphics Inc. Sec. Litig. , 183 F.3d 970, 987 (9th Cir.1999) )). When stock is sold at a profit but the price "generally continue[s] to climb," that suggests that "the timing of the...sales was not unusually prescient" and does not enhance the strength of the scienter inference. See Southland , 365 F.3d at 369.

Karnes sold KiOR stock between March 2012 and September 2013 and made $133,542 in profit. (Docket Entry No. 86 at ¶ 192). The allegedly false or misleading statements were made in May and August 2013. After making these statements, Karnes first sold stock on September 4, 2013, when the price was $1.93 per share. That is the lowest value at which Karnes sold. He sold 2,053 shares, or 1.84 percent of his ownership. This sale does not support an inference that Karnes made a false or misleading statement to cause a spike in the share price in order to take advantage of the increase by selling his stock. See Southland , 365 F.3d at 369. Nor did Karnes sell the largest percentage of his shares when the stock price peaked or was at a higher level. See Shaw , 537 F.3d at 543. And the SEC filings, referred to in the third amended complaint, show that Karnes increased his KiOR stock holdings over the class period. (See Docket Entry No. 92, Ex. 1 at p. 8). When a defendant "increase[s] his holdings of [company] stock...during the [c]lass period," it "negates any inference [of scienter]." Fire & Police Pension Ass'n of Colo. v. Abiomed, Inc. , 778 F.3d 228, 246 (1st Cir.2015).

B. Raising Capital

The plaintiffs allege that Karnes was motivated to make false statements about KiOR to raise capital and generally improve the company's finances. (Docket Entry No. 86 at ¶¶ 173–77). "The desire to raise capital in the normal course of business does not support a strong inference of scienter because virtually all corporate insiders share this goal." Owens , 789 F.3d at 539 ; see also Abrams , 292 F.3d at 434. These allegations do not support a strong inference of scienter.

C. Executive Compensation

The plaintiffs allege that Karnes was motivated to make false and misleading statements out of a desire to continue receiving his "lucrative" salary and compensation. (Docket Entry No. 86 at ¶ 178). Courts have rejected this argument as proving too much. "Incentive compensation can hardly be the basis on which an allegation of fraud is predicated." Abrams , 292 F.3d at 434 (quotation marks omitted). "[W]ere the opposite true, the executives of virtually every corporation in the United States would be subject to fraud allegations." Id. (quotation marks omitted); see also Fin. Acquisition Partners LP v. Blackwell , 440 F.3d 278, 290 (5th Cir.2006) ("Plaintiffs' mere allegation that the Individual Defendants were motivated by a desire to retain their jobs does not satisfy the scienter requirement.").

vii. Whether the Allegations Collectively Give Rise to a Strong Inference of Scienter

Although "[a] district court may best make sense of scienter allegations by first looking to the contribution of each individual allegation to a strong inference of scienter," it must "follow this initial step with a holistic look at all the scienter allegations." Owens , 789 F.3d at 537. The question is whether "all of the facts alleged, taken collectively, give rise to a strong inference of scienter." Tellabs I , 551 U.S. at 323, 127 S.Ct. 2499. "[T]he court must take into account plausible opposing inferences," including "plausible, nonculpable explanations for the defendant's conduct, as well as inferences favoring the plaintiff." Id. at 323–24, 127 S.Ct. 2499. "The strength of an inference cannot be decided in a vacuum. The inquiry is inherently comparative: How likely is it that one conclusion, as compared to others, follows from the underlying facts?" Id. at 323, 127 S.Ct. 2499.

At most, the third amended complaint supports an inference that Karnes was negligent in making the challenged May and August 2013 statements. There is no support for a plausible—much less strong—inference that Karnes was severely reckless in disregarding known or obvious facts about the state of KiOR's technology when he made the challenged statements. There is instead a strong, plausible, nonculpable inference that Karnes relied on data KiOR released—data the plaintiffs do not challenge as false or misleading—showing that the plant's yields and run-time percentages were improving. And although the third amended complaint allegations, if proven, could show that Karnes knew about the company's increasing debt, losses, and costs, the plaintiffs do not allege that the statements Karnes made disclosing those increases were false or misleading. Instead, the record shows that Karnes truthfully disclosed those facts on earnings and investor calls.

To survive dismissal under the PSLRA, the scienter inference "must be more than merely ‘reasonable’ or ‘permissible’—it must be cogent and compelling." Id. at 325, 127 S.Ct. 2499. The third amended complaint, taken as a whole, does not give rise to a cogent and compelling inference that Karnes acted with intent or severely reckless disregard of the likelihood of misleading investors when he made the May and August 2013 statements.

b. Cannon

The court may assess the scienter allegations holistically without first reviewing each allegation individually. See Owens , 789 F.3d at 536–37. Because some scienter allegations as to Cannon and Karnes overlap, and because the court has engaged in both an individual and a holistic review of the scienter allegations against Karnes, a holistic review is applied to the allegations against Cannon. The review gives special attention to the allegations unique to Cannon, including the allegations about the information the confidential witnesses provided to him. The "core-business" exception is also analyzed.

The factual allegations in the third amended complaint support the inference that Cannon misled investors by downplaying the technological problems at the Columbus plant as normal start-up issues, while simultaneously assuring investors that the technology was sound and that the plant was operating at a commercial scale. The allegations, if proven, plausibly show that Cannon misled investors to believe that the Columbus plant was a viable operation that had faced minor setbacks, overcome them, and was prepared for commercially viable production, both in the present and over the long haul. The question is whether the third amended complaint allegations give rise to a strong inference that Cannon made these statements with at least severely reckless disregard for their misleading nature.

Like Karnes, Cannon is alleged to have received reports from CW 6 and Paul O'Connor about lower-than-expected yields at the Houston demonstration unit. But unlike Karnes, Cannon made some of the statements the plaintiffs challenge very shortly after the Columbus plant started to operate, when data from the Houston demonstration unit was the only historical and objective evidence of how the Columbus plant would run. This is particularly true given the first-of-a-kind nature of the plant and the technology.

The allegations against Cannon are also more specific than those against Karnes. They include reports from CW 6 that Cannon met with William Coates, the COO, about the low yields at the plant, and that Coates relayed that specific information to Cannon. When Coates did, he was "fired and/or resigned immediately." (Docket Entry No. 86 at ¶ 52). The allegations include that in short order, O'Connor confirmed to Cannon what Coates and CW 6 had said. In March 2012, after an internal audit, O'Connor gave Kaul and Khosla his final report reflecting much lower yield capabilities at the plant. Kaul relayed that information to Cannon who, in reply, advised Kaul that O'Connor was being too pessimistic. (Id. at ¶ 72).

O'Connor and CW 6 occupied leadership roles in the company. Together, they had developed KiOR's technology. (Id. at ¶ 48). The allegations about the information they gave Cannon strongly support the inference that Cannon was reckless in disregarding their concerns and in firing or in forcing the resignation of another highly placed employee who pursued them, William Coates. These allegations also strongly support the inference that Cannon recklessly misled investors in representing, shortly after the Columbus plant began to operate, that KiOR was a viable business that had "proved" its technology could work on a commercial scale and that the issues facing the plant were merely temporary "start-up" glitches. Cannon made these representations even though the creators of KiOR's technology had told him, based in part on the test runs at the Houston demonstration unit, that the yields were, and the predicted production yields would be, much lower than reported. When Cannon heard this from the COO, he fired him or forced him to resign. Taken together, these allegations strongly support an inference that Cannon acted with scienter.

The strength of the scienter inference is enhanced by the allegations that after the Columbus plant had been running for several months, O'Connor and CW 6 again told Cannon that yields were lower than those Cannon was reporting and that his public statements were misleading. When O'Connor raised these problems with Cannon in May 2013, Cannon dismissed his concerns, telling O'Connor that he was being too pessimistic and that research and development had made "tremendous" progress. (Id. at ¶¶ 66, 73). When CW 6 met with Cannon and the vice-president, Christopher Artzer, several months later, in September 2013, it was to reiterate similar concerns. When CW 6 threatened to go to the board, Cannon allegedly asked CW 6 to sign a separation agreement. (Id. at ¶ 54). These allegations, taken as true, strongly support an inference that Cannon acted recklessly in disregarding information from the creators of KiOR's technology that his statements about the Columbus plant's production and progress were misleading investors. The severity of that recklessness is supported by Cannon's pattern of firing employees who came forward with concerns that his statements were misleading, or pressuring those employees to resign.

Although there are no allegations that Cannon actually knew about the severity of the technological problems—CW 6 is the only confidential informant alleged to have spoken with Cannon directly—the "core-business" exception also supports the inference that Cannon acted with scienter. The Columbus plant was KiOR's only plant. Its success on a commercial scale determined whether KiOR thrived or failed. Given the centrality of the plant's success to KiOR's business model, Cannon's role as the company CEO, and allegations from the company's senior scientist and chief technology officer that they directly and repeatedly communicated specific concerns to Cannon that the public representations were misleading, there is an inference to be drawn under the "core-business" exception that Cannon recklessly disregarded the signs and reports that KiOR faced increasingly difficult problems with its technology as 2013 progressed. Indeed, Cannon's own statements on earnings calls routinely disclosed problems at the Columbus plant, only to downplay them by telling investors that they were minor, typical of a start-up phase, and temporary.

Taken together, the allegations are similar to those in other cases in which the Fifth Circuit has applied the "core-business" exception to hold that a CEO or president, by virtue of his position and in light of the surrounding circumstances, acted with scienter. See Dorsey , 540 F.3d at 342–43 (company was "a one-trick pony" without an "extensive operating [history]" whose success was "dependent" on the facts underlying the basis of the statement alleged to be false or misleading); Plotkin , 407 F.3d at 700 ("reasonable to assume, given the importance of the[ ] deals to the company," that company "would have familiarized itself" with information allegedly withheld); Nathenson , 267 F.3d at 424–25 (one-product company with its future prospects dependent on the product about which the defendants had made statements alleged to be false or misleading). And although KiOR had 183 employees by the end of fiscal year 2013, making it substantially larger than the companies to which the Fifth Circuit has applied the "core-business" exception in other cases, there is no bright-line rule based on the company's size. To the contrary, "[t]he ‘special circumstances' cases exhibit some combination of four considerations that might tip the scales in favor of an inference of scienter," including not only the size of the company, but also, as here, when the statements at issue concern transactions "critical to the company's continued vitality" and when "the misrepresented or omitted information at issue would have been readily apparent to the speaker." Diodes , 810 F.3d at 959 (emphasis added). No one "special circumstance" is dispositive.

The allegations more strongly support an inference of Cannon's culpable scienter as the class period progressed. The inference is perhaps at its strongest by January 2014, when Cannon disclosed the severity of the problems facing the plant but assured investors that operations could be salvaged. A month later, O'Connor concluded "that the main problems at Columbus [were] already discernable in the [Houston demonstration unit] and [were] therefore structural and not ‘just’ operational issues." (Docket Entry No. 86 at ¶ 67). Two months later, the plant was shut down permanently. See Plotkin , 407 F.3d at 698.

The competing, nonculpable explanation for Cannon's statements, like Karnes's, is that he relied on data showing increasing yields and run-time percentages at the plant. The difference here is that the inference that Cannon acted with scienter is at least as strong as the inference that he acted without it. "A complaint will survive...only if a reasonable person would deem the inference of scienter cogent and at least as compelling as any opposing inference one could draw from the facts allege." Tellabs I , 551 U.S. at 324, 127 S.Ct. 2499. "[A] tie favors the plaintiff." Lormand , 565 F.3d at 254.

4. Summary of the Conclusions

The specified statements about the Columbus plant are not protected under the safe harbor. The third amended complaint plausibly alleges with particularity that those statements were misleading. The third amended complaint does not give rise to a strong inference that Karnes made the May and August 2013 statements with scienter. The third amended complaint does give rise to a strong inference that Cannon made the statements from November 2012 to January 2014 with scienter.

B. The Oral Statements About Future Yields

The plaintiffs allege that either Cannon or Karnes made the following misleading predictions about future yields:

• On a November 2012 earnings call, Karnes stated: "[W]e do think our prior guidance of around 500,000 to one million gallons of blend stock sales, seems like a reasonable planning [for the fourth quarter of 2012] assuming the start-up continues as we've got it planned...." (Docket Entry No. 86 at ¶ 102).

• On a March 2013 earnings call, Cannon stated: "[O]ur focus at Columbus turns toward achieving steady state operations. Until we have the plant fully lined down, which we believe could take at least nine months, any guidance we might give to the basic metrics of plant performance that we use in the petrochemicals or have to have an unusually wide error bar around that. Given what I just said about the fact that production volumes in the first quarter will be negligible, looking at the remaining nine months of the year, we believe our volume expectations to be approximately 3 to 5 million gallons of the balance of the year." (Id. at ¶ 108).

• On a March 2013 earnings call, Karnes stated: "[F]or planning purposes we are assuming that our 2013 total production target of 3 to 5 million gallons will simply occur linearly starting from essentially zero in [the first quarter]." (Id. at ¶ 109).

• On a May 2013 earnings call, Cannon stated: "Consistent with our previous guidance, we expect that total fuel production during the second quarter will range between 300,000 and 500,000 gallons, keeping us on track to fall within our projected production range of 3 million to 5 million gallons for 2013." (Id. at ¶ 117).

• On a May 2013 earnings call, Karnes stated: "While our limited production history makes forecasting difficult, we don't see any reason at this point to change our prior production guidance as [Cannon] mentioned, of 3 million to 5 million gallons for the year. Without any more operational data, for internal purposes, we're just assuming this production ramp occurs fairly linearly from 5,000 gallons in [the first quarter] to between 300,000 and 500,000 gallons in [the second quarter], as [Cannon] mentioned. While we expect shipments to pick up going forward, shutdowns will continue to be par for the course until the plant fully stabilizes, so shipments can be expected to be sporadic." (Id. at ¶ 118).

1. Whether the Projections Were Accompanied by Meaningful Cautions

These statements are forward looking and accompanied by cautionary statements. The plaintiffs argue that the safe harbor does not apply because the cautions were "drowned out" by other false or misleading statements Cannon or Karnes made and were not "meaningful" because the statements referred investors to boilerplate language in KiOR's SEC filings.

The November 2012 and March and May 2013 earnings calls identified the challenged projections as forward looking and accompanied them with statements warning that actual results might materially differ. See 15 U.S.C. § 78u–5(c)(2)(A) ; Emp'rs Teamsters Local Nos. 175 & 505 Pension Trust Fund v. Clorox Co. , 353 F.3d 1125, 1132–33 (9th Cir.2004) (starting conference call with statement that "I need to remind you that any forward-looking statements made on this call represent our best judgment as to what may occur in the future," that "[t]he Company's actual results will depend on a number of competitive and economic factors, some of which may be outside the control of the Company," and that referred listeners to meaningful cautionary language in the company's Form 10-K made the warnings meaningful and therefore protected the forward-looking statements under the safe harbor).

KiOR's finance director, Max Kricorian, began each earnings call with the following caution:

I would like to remind everyone that statements will be made during this call that are not historical facts and are forward-looking statements. These statements about the company's future expectations, plans and prospects include statements regarding the timing for production from the company's Columbus plant. The company's plan to build it next commercial plant in Natchez, Mississippi. Achievement of advances in our technology platform are positive sign, including improvements over yield, the accuracy of our estimates regarding expenses, construction costs, future revenue and capital requirements and other statements containing the words, believes, anticipates, plans, expects, intends, and similar expressions. These forward looking statements are based on current expectations and assumptions that are subject to risks and uncertainties.

The company cautions that a number of important factors could cause our actual future results and other future circumstances to differ materially from those expressed in such forward-looking statements.

(Docket Entry No. 94, Ex. 3 at p. 513–14, 577–78, 729–30).

Cannon and Karnes supplemented this introduction with more specific warnings that the projections might not materialize. In the November 2012 earnings call, Cannon stated that the yield projection was "reasonable planning" but depended on "the start-up continu[ing]" to perform as "planned." During the March 2013 call, Cannon warned that the yield projections had "an unusually wide error bar" because the Columbus plant had not yet reached steady-state production. Karnes also stated that these yield projections were for "planning" purposes only. During the May 2013 call, Karnes noted that forecasting yields was "difficult" and that the projections assumed linear production growth "for internal purposes." Shutdowns, Karnes told investors, would also be "par for the course."

Cannon and Karnes confirmed these oral warnings in writing. In the calls, they referred investors to "additional information concerning factors that could cause actual results to materially differ" that were "contained in a readily available written document," and they identified the specific document during the call. See 15 U.S.C. § 78u–5(c)(2)(B)(i)–(ii). Each earnings call directed investors to KiOR's SEC filings. These filings are clearly "readily available written document[s]" under the safe harbor. See Southland , 365 F.3d at 372 (citing 15 U.S.C. § 78u–5(c)(3) ).

In the November 2012 earnings call, KiOR's finance director told investors where in the relevant Forms 10-K and 10- Q they could find written information about the risks of investing:

These important factors, and other factors that could potentially affect the company's financial results are described in the section called risk factors, in the company's most recent annual report on Form 10-K for the fiscal year ended December 31, 2011 as updated by the company's quarterly report from Form 10-Q, for the quarter ended June 30, 2012, and the company's other filings with the SEC, which are available through the investor relations section of the KiOR website at www.kior.com, or on the SEC website at www.sec.gov.

The company may change its intentions, belief or expectations at any time, and without notice, based upon any changes in such factors and the company's assumptions or otherwise. The company undertakes no obligation to release publicly any revisions to any forward-looking statements will affect events or particular circumstances after today or through respective currents of anticipated events. Therefore, you should not rely on these forward-looking statements as representing the company's views as of any day subsequent to today.

(Docket Entry No. 94, Ex. 3 at p. 514).

In the March 2013 earnings call, KiOR's finance director again told investors where to find written risk statements:

The company cautions that a number of important factors could cause our actual future results and other future circumstances to differ materially from those expressed in any such forward-looking statements. These important factors, and other factors that could potentially affect the company's financial results are described in the section called risk factors, in the company's most recent annual report on Form 10-K for the fiscal year ended December 31, 2011 and in the company's other quarterly filing with the SEC, which are available through the Investor Relations section of the

KiOR website at www.kior.com, or on the SEC website at www.sec.gov.

The company may change its intentions, beliefs or expectations at any time, and without notice, based upon any changes in such factors and the company's assumptions or otherwise. The company undertakes no obligation to release publicly any revisions to any forward-looking statements to reflect events or circumstances after today or through respective currents of anticipated—and anticipated events. Therefore, you should not rely on these forward-looking statements as representing the company's views as of any day subsequent to today March 18, 2013.

(Docket Entry No. 94, Ex. 3 at p. 578).

In the May 2013 earnings call, KiOR's finance director told investors:

The company cautions that a number of important factors could cause our actual future results and other future circumstances to differ materially from those expressed in any such forward-looking statements. These important factors and other factors that could potentially affect the company's financial results are described in the section called Risk Factors in the company's most recent annual report on Form 10-K for the fiscal year ended December 31, 2012 and in the company's other quarterly filings with the SEC, which are available through the Investor Relations section of the KiOR website at www.kior.com, or on the SEC website at www.sec.gov.

The company may change its intentions, beliefs or expectations at any time, and without notice, based upon any changes in such factors and the company's assumptions or otherwise. The company undertakes no obligation to release publicly

any revisions to any forward-looking statements to reflect events or circumstances after today or to reflect the occurrence of unanticipated events. Therefore, you should not rely on these forward-looking statements as representing the company's views as of any date subsequent to today[,] May 9, 2013.

(Docket Entry No. 94, Ex. 3 at p. 730).

The parties dispute whether "the information contained in [the] written document[s] [referenced orally]"—the SEC filings—satisfied the requirement that the cautionary language be meaningful. See 15 U.S.C. § 78u–5(c)(2)(B)(iii) ; id. § 78u–5(c)(1)(A). The 2011 Form 10-K had 14 pages of disclosed risks. (Docket Entry No. 94, Ex. 3 at p. 386–99). It identified specific risks relevant to the yield projections, including the following:

We are a development stage company and have not generated any revenue, and our business will not succeed if we are unable to commercialize successfully our cellulosic gasoline and diesel.

We are a development stage company with a limited operating history, and we have not yet commercialized our cellulosic gasoline and diesel nor have we generated any revenue. We are subject to the substantial risk of failure facing businesses seeking to develop new products. Certain factors that could, alone or in combination, prevent us from successfully commercializing our products include: technical challenges developing our commercial production processes that we are unable to overcome; ...[and] our ability to achieve commercial-scale production of cellulosic gasoline and diesel on a cost-effective basis and in the time frame we anticipate.....

...

We have no experience producing cellulosic gasoline and diesel at the scale needed for the development of our business or in building the facilities necessary for such production, and we will not succeed if we cannot effectively scale our proprietary technology platform and process design.

We must demonstrate our ability to apply our proprietary technology platform and process design at commercial scale to convert biomass into cellulosic gasoline and diesel on an economically viable basis. Such production will require that our proprietary technology platform and process design be scalable from our demonstration unit to commercial production facilities. We have not yet completed construction of or operated a commercial-scale production plant, and our technology may not perform as expected when applied at the scale that we plan or we may encounter operational challenges for which we are unable to devise a workable solution. In particular, our initial-scale commercial production plant under construction in Columbus, Mississippi is a first-of-kind project, and we cannot assure you that it will be completed on the schedule that we intend or at all. If and when completed, our initial-scale commercial production plant may not process biomass at designed levels or produce our cellulosic gasoline and diesel at acceptable yields, and we may be unable to improve its performance. As a result of these risks, we may be unable to achieve commercial-scale production in a timely manner, or at all. If these risks materialize, our business and ability to commercialize our cellulosic gasoline and diesel would be materially and adversely affected.

...

We may be unable to realize expected economies of scale, reduce our feedstock costs, increase our overall yields and optimize the composition of our renewable transportation fuels, which could limit our ability to sell our products at

competitive prices and materially and adversely affect our business and prospects.

We may be unable to realize expected economies of scale, reduce our feedstock costs, increase our overall yields and optimize the composition of our renewable crude oil in order to produce our renewable fuel products on a cost competitive basis with existing petroleum-based fuel products without government incentives. In particular, we may be unsuccessful in incorporating lower grade woody biomass, such as logging residues, branches and bark, in our process to reduce our feedstock costs or maintain our yields. In addition, our research and development efforts may fail to increase the yield of our [biomass fluid-catalytic-cracking] process such that we may be unable to produce renewable transportation fuels at the costs or in the quantities that we anticipate. Our failure to achieve these efficiencies or improvements over time could limit our ability sell our products at competitive prices and materially and adversely affect our business and prospects.

(Docket Entry No. 94, Ex. 3 at p. 386–87, 390).

The June 2012 Form 10-Q stated that "[t]here have been no material changes from the risk factors disclosed in our Annual Report on Form 10-K for the year ended December 31, 2011," but noted specific exceptions requiring updated information under some of the subheadings set out in the 2011 Form 10-K. (Docket Entry No. 94, Ex. 3 at p. 489–91). The 2012 Form 10-K identified the same risk factors set out in the 2011 Form 10-K, but again updated the information. (Docket Entry No. 97, Ex. 2 at p. 67–93).

The Forms 10-K and 10-Q included specific reasons the actual results could be materially different from those predicted in the forward-looking statements, including specific reasons why the Columbus plant's yields might be lower than anticipated. The warnings specifically noted that KiOR was trying to produce a new product that had a "substantial risk of failure." KiOR had a "limited operating history" and had not "yet commercialized [its] cellulosic gasoline and diesel." The warnings identified specific challenges, including "technical challenges developing... commercial production processes" and achieving "commercial-scale production of cellulosic gasoline and diesel on a cost-effective basis." The warnings noted that bringing KiOR's technology to commercial scale could see the company's "research and development efforts... fail to increase the yield of [the] [biomass fluid-catalytic-cracking] process" and the inability "to produce renewable transportation fuels at the costs or in the quantities that [it] anticipate[d]."

These disclosures are different from warnings the case law has rejected as "boilerplate" and not meaningful. Compare with Lormand , 565 F.3d at 244 ("[The statements are] not guarantees of future performance...and involve known and unknown risks and other factors that could cause actual results to be materially different from any future results expressed or implied by them."), and Plotkin , 407 F.3d at 694 ("These forward-looking statements involve numerous risks, uncertainties and assumptions, and actual results could differ materially from anticipated results."). The Forms 10-K and 10-Q included "company-specific warnings based on a realistic description of the risks applicable to the particular circumstances," see Southland , 365 F.3d at 372, rather than "limited, general, and vague" risks, see Lormand , 565 F.3d at 246.

The plaintiffs argue that these risk disclosures were not meaningfully cautionary because the same language appeared in both the 2011 and the 2012 Form 10-Ks. "Had the [defendants] intended the language to amount to something more than generic, boilerplate material, they would have modified it [in 2012] to account for the ‘start-up’ issues they encountered during the ‘start-up’ phase following mechanical completion of the plant in September 2012." (Docket Entry No. 97 at p. 27).

The record shows, however, that the 2012 Form 10-K was revised to reflect changes in the risks KiOR faced. (Docket Entry No. 99, Ex. A). The 2012 Form 10-K added the following language:

We are vulnerable to disruptions to our renewable fuel production operations, because all our production operations exist in a single first-of-kind plant.

Because all of our production operations currently exist in a single plant in Columbus, Mississippi, significant and prolonged disruptions at the plant would have a material adverse effect on our business, financial condition and results of operations. Our initial-scale commercial production plant in Columbus, Mississippi is our only existing production plant. We are in the start-up phase at this first-of-kind plant, and only commenced limited sales of our cellulose diesel in March 2013. During the Columbus plant's early years of production and until the plant stabilizes operationally, the plant may be more susceptible to start-ups and shutdowns that will disrupt and delay our production activities for significant time periods. We do not know when or if we will achieve stable operations at Columbus. Even if we achieve steady-state operations at the plant, we will be required to perform periodic maintenance and turnarounds at the plant. All or a significant portion of the plant may be disrupted during maintenance or a turnaround. These disruptions will make any future revenue and cash flows unpredictable.

(Docket Entry No. 99, Ex. 1 at p. 8).

This new language added to the 2012 Form 10-K specifically warned that the Columbus plant was in the "start-up phase" and had only "limited sales"; that production had not stabilized; and that the plant was "more susceptible" to production delays for "significant time periods," causing "prolonged disruptions." The 2012 Form 10-K provided investors with updated risk information in tailored warnings identifying problems that could prevent KiOR from realizing its yield projections. The revised and new disclosure undercuts the plaintiffs' argument that as the Columbus plant confronted increasingly complex technological challenges, KiOR's risk disclosures remained the same.

The case law recognizes that the repetitious use of boilerplate warnings weighs against finding them meaningful cautions. But the case law does not prohibit using similar warnings from past years that continue to apply or reject that use as "boilerplate." To the contrary, courts have found language meaningfully cautionary even when the statements repeated earlier disclosures. See, e.g. , In re Alamosa Holdings, Inc. , 382 F.Supp.2d 832, 845 (N.D.Tex.2005) ("[I]nvestors were put on notice time and again that changes could cause significant decline in [the defendant's] attraction of new subscribers."); In re Michaels Stores, Inc. Sec. Litig. , No. 03–cv–0246, 2004 WL 2851782, at *7 (N.D.Tex. Dec. 10, 2004) (press releases containing "the same cautionary language" were meaningful and protected under the safe harbor).

In Lormand , the Fifth Circuit identified three factors to consider in determining whether a disclosure has meaningful cautionary language: whether the disclosure is "generic and formulaic"; whether the language used is "boilerplate" rather than substantively meaningful; and whether "each alleged misrepresentation" has the same disclosure "only with slight variations." 565 F.3d at 245. The 2011 and 2012 Form 10-Ks had some similar language, but the 2012 Form 10-K and the oral statements made on each earnings call also contained new and specific information about factors affecting KiOR's yield projections. The 2012 Forms 10-K included warnings that the Columbus plant had not reached steady-state operations and was vulnerable to production shutdowns for significant periods that could cause significant disruptions. The earnings-call statements confirmed that these events were happening at the Columbus plant, warned that the projections were for planning purposes only, and disclosed the linear-growth assumptions underlying them while noting that linear growth had not been realized. These are meaningful cautionary disclosures.

The plaintiffs have not cited a case holding that reusing some warning statements from a year-old SEC filing, along with revised and updated information, means that the statements cannot be meaningfully cautionary. Lormand undermines the argument by requiring courts to examine the context and the content of disclosures to determine whether they are meaningful. See id. And the record does not support the plaintiffs' argument that KiOR simply recycled the 2011 Form 10-K risk disclosures in its 2012 Form 10-K, as the differences between the two documents show. Compare with id. at 247 (company failed to disclose "certain dangers that had already begun to materialize"), and In re Harman , 791 F.3d at 107 ("[T]he Company's cautionary statements remained unchanged despite a significant change in circumstances of material importance to an investor.").

The plaintiffs cite Asher v. Baxter Int'l, Inc. , 377 F.3d 727 (7th Cir.2004), in support of their argument, but that case is different. In Asher , the problem the court emphasized was that the record provided no basis "to conclude that [the defendant] mentioned those sources of variance that (at the time of the projection) were the principal or important risks." Id . at 734. As a result, it was unclear whether "the major risks [the defendant] objectively faced when it made its forecasts were exactly those that, according to the complaint, came to pass, [even though] the cautionary statement mentioned none of them." Id. This case is, if anything, Asher 's mirror image. The plaintiffs' allegations of the problems KiOR ended up facing were disclosed. Those risks included failing research-and-development efforts to increase yields, faltering technology that did not and could not perform on the needed and predicted commercial scale, and a malfunctioning plant that experienced prolonged, significant shutdowns. See In re Alamosa Holdings , 382 F.Supp.2d at 845 ("[T]he very risks Plaintiffs are complaining of were expressly contained in the...cautionary language."). And in Asher , unlike the present case, "the cautionary language remained fixed even as the risks changed." 377 F.3d at 734. In this case, by contrast, both the Forms 10-K and 10-Q and the statements made during the earnings calls updated the risk disclosures, "point[ing] to the principal contingencies that could cause actual results to depart from the projection." Compare with id.

The plaintiffs also argue that the disclosures were not meaningful because misleadingly rosy earnings-call statements "drowned out" the warnings. The plaintiffs cite statements in Lormand that when cautionary disclosures "fail[ ] to correct the false impression created by the defendants' statements or to supply the truth that they omitted," and "reasonable minds could disagree as to whether the mix of information in the allegedly actionable document is misleading, the statutory safe harbor provision cannot provide the basis for dismissal as a matter of law." 565 F.3d at 247–48 (alterations omitted) (quotation marks omitted). These statements were dicta in Lormand because the court had already found the safe harbor inapplicable on other grounds. See note 1. The court then addressed in the alternative whether a cautionary statement was itself misleading because it did not correct other allegedly false or misleading statements. See Lormand , 565 F.3d at 247–48.

The issue in Lormand was different. The PSLRA safe harbor requires cautionary language that "identif[ies] important factors that could cause actual results to differ materially from those in the forward-looking statement." 15 U.S.C. § 78u–5(c)(1)(A)(i). Lormand is one of many cases recognizing that cautionary disclosures that are themselves misleading cannot be "meaningful" and are not protected under the safe harbor. The cautionary statements in Lormand not only failed to "disclose the specific risks and their magnitude," but also characterized "certain dangers that had already begun to materialize" as only possible future risks, not as present problems or clearly present risks. 565 F.3d at 247. An issuer cannot "list its lines of business, say ‘we could have problems in any of these,’ and avoid liability for statements implying that no such problems were on the horizon even if a precipice was in sight." Asher , 377 F.3d at 733 ; see also In re Harman , 791 F.3d at 106 ("The cautionary language...fails to account for the materialization, rather than abstract possibility, of the important risk posed...."); Rombach v. Chang , 355 F.3d 164, 173 (2d Cir.2004) (the safe harbor would not protect "someone who warns his hiking companion to walk slowly because there might be a ditch ahead when he knows with near certainty that the Grand Canyon lies one foot away."). Lormand held that the plaintiffs had adequately alleged that even though cautionary language warned of future risks, it did not "provide sufficiently meaningful caution about clearly present danger that was materializing." 565 F.3d at 247.

It is unclear whether the plaintiffs intend to assert this argument. If so, the factual allegations and the record provide no support. Cannon orally disclosed the technical problems the confidential witnesses identified. Besides oral statements disclosing risks, during the earnings calls in November 2012 and March and May 2013, investors were also referred to written risk disclosures in SEC filings that were updated to account for changes in the problems KiOR encountered after the plant began operating in late 2012 and continued operations—with interruptions—into early 2013. The disclosures the plaintiffs allege, and the materials properly considered in this Rule 12(b)(6) motion, clearly distinguish this case from Lormand , in which the cautionary statements failed to disclose "certain dangers that had already begun to materialize." 565 F.3d at 247.

Although the court has found that the allegations of Cannon's and Karnes's representations about the dangers to, and the viability of, the Columbus plant sufficiently pleaded their misleading nature, that does not preclude the conclusion that the yield-projection risk warnings were meaningful. The projections, unlike the misleading statements, concerned only the yield, and the oral and written risk disclosures gave investors updated and specific information about the limits of, and the bases for, the projections, as the safe harbor demands.

The November 2012 and March and May 2013 earnings-call yield projections are protected under the safe harbor. 2. Whether the Projections Were Made with "Actual Knowledge"

Alternatively, the safe harbor protects these statements because the plaintiffs have not alleged that Cannon or Karnes made them with "actual knowledge" that they were false or misleading. "Since [the safe harbor] specifies an ‘actual knowledge’ standard, the scienter requirement for forward-looking statements is stricter than that for statements of current fact. Whereas liability for the latter requires a showing of either knowing falsity or recklessness, liability for the former attaches only upon proof of knowing falsity." Avaya , 564 F.3d at 274 (collecting cases).

As discussed, the plaintiffs have not plausibly alleged a strong inference of scienter as to Karnes, including actual knowledge. As for Cannon, the court concluded that even though the plaintiffs had not alleged the precise means by which he learned of the severity of the problems facing the Columbus plant, the factual allegations gave rise to a strong inference of severe recklessness, an inference that was at least as cogent and compelling as competing, nonculpable explanations. "In the case of the forward-looking statements, however, an inference of recklessness does not avail plaintiffs—that is, it must be placed on the nonculpable-explanation side of the balance when we weigh competing inferences." Id. The factual allegations in the third amended complaint do not support a strong inference that Cannon actually knew that the yield projections were false or misleading. The safe harbor's "actual knowledge" prong provides an independent ground for dismissing the claims based on the yield projections themselves.

C. The Written Statements About Future Yields

The plaintiffs allege that the following statements about future yields were false or misleading:

• In a September 2013 press release, Cannon stated: "With the [biomass fluid-catalytic-cracking] section of the Columbus plant currently producing additional oil, we believe that we are well-positioned to build on the progress made during July and August and to produce additional volumes of cellulosic fuel for American vehicles consistent with our most recent guidance [of 1 to 2 million gallons]." (Id. at ¶ 136).

• In a November 2013 press release, Cannon stated: "I am happy to report again that we are seeing significant operational progress at our Columbus plant and we believe that we are turning the corner toward steady-state operations....As a result, we believe that with stable production over the balance of the year, our full year production levels will exceed 1 million gallons." (Id. at ¶ 138).

Cannon argues that these statements are immaterial puffery. The plaintiffs respond that although an isolated promise of progress may be immaterial, the promises here were not only repeated, but also became more and more misleading and material over time.

In isolation, Cannon's statements that KiOR was "well positioned to build on the progress made during July and August," and that "we are seeing significant operational progress at our Columbus plant and we believe that we are turning the corner toward steady-state operations" are vague assertions of corporate optimism. "[I]t is well-established that generalized positive statements about a company's progress are not a basis for liability." Nathenson , 267 F.3d at 419. A press release stating that a company is " ‘making steady progress' is precisely the sort of generalized positive characterization that is not actionable under the securities laws." Rosenzweig , 332 F.3d at 870 ; see also Southland , 365 F.3d at 372 ("The generalized, positive statements about the company's competitive strengths, experienced management, and future prospects are not actionable because they are immaterial.").

Some of Cannon's promises of progress, however, were tethered to statements that the Columbus plant was "currently producing additional oil" and "turning the corner toward steady-state operations." These factual statements about current operations were a basis for projecting that the Columbus plant would produce 1 to 2 million gallons in 2013. Cannon did not simply state that KiOR was generally "well positioned" or ready to "turn the corner." Rather, Cannon's specific and concrete statements about KiOR's progress toward million-gallon yields described objectively verifiable, current facts.

In re BP is similar and instructive. BP issued a report to investors describing recommended safety measures and stating that it had made progress implementing the recommendations. In re BP , 843 F.Supp.2d at 758–59. The question was whether BP's statements that it was making progress were misleading because BP was not following the specific "roadmap" it had released. Id. at 759. BP's repeated promises of progress could be "measured against" the publicly available document and were therefore not immaterial puffery. Similarly, Cannon's promises of progress could be measured against KiOR's own "roadmap": its yield projections. The promises were not so vague, subjective, or general that reasonable investors would find them unimportant.

For similar reasons, aspects of these September and November 2013 press-release statements are not forward looking. A " ‘mixed present/future statement is not entitled to the safe harbor with respect to the part of the statement that refers to the present.’ " Spitzberg , 758 F.3d at 691 (quoting Tellabs II , 513 F.3d at 705 ). " ‘The mere fact that a statement contains some reference to a projection of future events cannot sensibly bring the statement within the safe harbor if the allegation of falsehood relates to non-forward-looking aspects of the statement.’ " Id. at 691 n. 25 (quoting In re Stone & Webster, Inc., Sec. Litig. , 414 F.3d 187, 213 (1st Cir.2005) ). In Tellabs II , for example, the statement that product sales were "still going strong" was not protected under the "safe harbor with regard to the statement's representation concerning current sales." 513 F.3d at 705. Similarly, in In re Stone & Webster , the statement that the company "has on hand and has access to sufficient sources of funds to meet its anticipated...needs" was not forward looking because "the alleged falsehood was in the fact that the statement claimed that the Company had access to ample cash at a time when the Company was suffering a dire cash shortage. The claim was not that the Company was understating its future cash needs." 414 F.3d at 207, 213. In Avaya , by contrast, statements that "we are on track to meet our [financial] goals for the year" and "[o]ur first quarter results position us to meet our goals for the year" were "assertions of current fact...too vague to be actionable," and could not "meaningfully be distinguished from the future projection of which they [were] a part." 564 F.3d at 247, 255. Unlike Tellabs II and Stone & Webster , the statements in Avaya did "not justify the financial projections in terms of any particular aspect of the company's current situation; they [said] only that, whatever that situation [was], it [made] the future projection attainable." Id. at 254–55. "Such an assertion is necessarily implicit in every future projection." Id. at 255.

Cannon's September 2013 statement that the Columbus plant was "currently producing" additional oil used this current fact to predict that KiOR was "well positioned" to make progress in achieving projected production yields. The alleged falsehood is that the plant was producing enough oil to meet the projected yields. This factual statement about the plant's current oil production is similar to the factual statements in Tellabs II about the strength of current sales and the statement in Stone & Webster about the current size of cash reserves. The statement is not forward looking. Cannon did not merely state that KiOR could meet its million-gallon-yield goal. He also "justif[ied] the...projection[ ] in terms of a[ ] particular aspect of the company's situation." See Avaya , 564 F.3d at 255. Cannon "advert[ed] to a particular current fact" instead of expressing only the company's "continuing comfort with the earlier...annual projection." See id. at 256.⁹

9 Compare with Part IV.B (discussing forward-looking statements reiterating the projections without referring to current facts).

Cannon's November 2013 statement is similar. He based the yield projection on representations about KiOR's current status, including that the company had made "significant operational progress" and was "turning the corner" to reach steady-state production. This statement did more than "reaffirm[ ] [the] projection" made earlier; it represented to investors the current state of operations at the Columbus plant, a representation that the plaintiffs allege was false or misleading. See id. The projection that "our full year production levels will exceed 1 million gallons," however, is clearly forward looking.

To the extent these statements are not forward looking, they are actionable for the reasons discussed above in Part IV.A of this opinion. To the extent these statements are forward-looking projections, they are protected under the safe harbor because the third amended complaint does not allege that Cannon made them with actual knowledge. See Part IV.B.2.

V. Control-Person Liability

Karnes is not liable for primary violations under § 10(b), so he cannot be liable under a control-person theory. "Control person liability is secondary only and cannot exist in the absence of a primary violation." Southland , 365 F.3d at 383.

Cannon argued that he could not be held liable under a control-person theory because the third amended complaint did not plausibly allege a primary violation. (Docket Entry No. 94 at p. 55). At this stage, there is no basis for dismissing the control-person claims against Cannon because the § 10(b) claims against him have not been dismissed.

The plaintiffs argue that Khosla can be held liable under a control-person theory because he had the ability to control the challenged statements. In support, they cite allegations in the third amended complaint that Khosla was the majority shareholder, appointed Samir Kaul to the board, and held the largest amount of secured debt. (Docket Entry No. 97 at p. 62–63). The problem is that the third amended complaint does not allege facts supporting a plausible inference that Khosla had the ability to control Cannon's specific, allegedly misleading public statements. Heck , 775 F.3d at 283. Assuming that Khosla was "deeply involved in and responsible for KiOR's operational strategy and financing" from "cradle to grave," (Docket Entry No. 97 at p. 63), without factual allegations showing or supporting an inference that he had the ability to control the public statements Cannon made as CEO, the control- person claims against Khosla must fail. They are dismissed.

VI. Conclusion

The claims against Karnes and Khosla are dismissed. Because this is the third amended complaint, further amendment would be futile, and the motions to dismiss filed by Karnes and Khosla are granted, with prejudice. (Docket Entry Nos. 92, 93).

The claims against Cannon are not dismissed. His motion to dismiss is denied. (Docket Entry No. 94). His statements from November 2012 to January 2014 about how the Columbus plant operated are actionable. See Part IV.A. Some of Cannon's statements cannot be the basis for liability, however. These statements include:

• The statements about the Houston demonstration unit, see Part III;

• The oral statements about future yields, see Part IV.B; and

• The written press-release statements about future yields, to the extent they are forward looking, see Part IV.C.

A status and scheduling conference is set for May 25, 2016 at 10:00 a.m. to set a scheduling order for discovery, subsequent motions, and trial.