Opinion
5:20-cv-05194-EJD
03-31-2023
IN RE INTEL CORP. SECURITIES LITIGATION
ORDER GRANTING MOTION TO DISMISS WITH LEAVE TO AMEND RE: ECF NO. 56
EDWARD J. DAVILA UNITED STATES DISTRICT JUDGE
Lead Plaintiffs KBC Asset Management NV and SEB Investment Management AB (collectively, “Lead Plaintiffs”) bring this putative class action against Defendants Intel Corporation (“Intel”), former Intel CEO Robert H. Swan, Intel CFO George S. Davis, and former Intel Chief Engineering Officer Dr. Venkata S.M. Renduchintala (collectively, the “Individual Defendants,” and with Intel, “Defendants”), alleging violations of §§ 10(b) and 20(a) of the Securities Exchange Act of 1934 and Rule 10b-5 promulgated thereunder. See Consolidated Compl. (“Compl.”), ECF No. 53. Lead Plaintiffs bring this action individually and on behalf of those who purchased or acquired Intel common stock from October 25, 2019 through October 23, 2020 (the “Class Period”). Id. at 1.
Davis was identified as a current officer of Intel at the time the Consolidated Complaint was filed.
Before the Court is Defendants' motion to dismiss under Federal Rules of Civil Procedure 8(a), 9(b), and 12(b)(6) and the Private Securities Litigation Reform Act of 1995 (“PSLRA”). Mot. to Dismiss (“Mot.”), ECF No. 56, at 1. The Court finds this matter suitable for decision without oral argument. Civil L.R. 7-1(b). Having considered the parties' submissions, the Court Grants Defendants' motion to dismiss with leave to amend.
BACKGROUND
For purposes of this motion to dismiss, the Court accepts as true the allegations of the Consolidated Complaint. Reese v. BP Exploration (Alaska) Inc., 643 F.3d 681, 690 (9th Cir. 2011).
Intel is a semiconductor company that designs and manufactures microprocessors and other semiconductor products for use in computers, data center servers, communications devices, and other digital electronic devices. Compl. ¶ 20, 26. It is headquartered in Santa Clara, California, and its stock trades on NASDAQ under the symbol “INTC.” Id. ¶ 20.
The semiconductor industry is marked by two features relevant to this case. First, semiconductor companies can broadly be characterized as occupying one of two roles: design or fabrication. Id. ¶ 32. Companies that focus only on chip design and marketing employ what is called a “fabless” business model because those companies do not own chip making facilities called “fabs.” Id. ¶¶ 29, 32. On the other hand, companies that only fabricate chips are called “foundries,” and fabless companies will send their chip designs to foundries to manufacture on their behalf. Id. ¶ 32. Some companies, however, perform both design and fabrication functions and are called “integrated device manufacturers,” or “IDMs.” Intel is (mostly) such an IDM. Id. ¶ 29. For its leading-edge chips-those based on the most current, advanced technology-Intel both designs and fabricates the chips in-house. Id. ¶¶ 29, 31. For older, trailing-edge chips and non-Intel designed chips obtained from acquisitions of other companies, Intel outsources some production to foundries. Id. ¶ 31. This business model allows Intel to realize efficiencies by avoiding intermediaries, to coordinate manufacturing capacity with demand, and to better safeguard its intellectual property by keeping its knowledge and expertise in-house. Id. ¶ 30.
The second key feature is that expectations and industry economics are closely tied to an assumption known as Moore's Law, which stems from an observation by Intel co-founder Gordon Moore. In the 1960s, Moore observed that every two years, the number of transistors that could be fabricated in a given area of silicon wafer would double, meaning that chips would grow smaller at a rapid rate. Id. ¶ 39. He fore casted that this trend would continue, and history has borne out his prediction. Id. ¶¶ 39, 41. Because development has kept pace with Moore's Law, the semiconductor industry and its analysts now expect chip sizes to shrink in accordance with the cadence of Moore's Law. Id. ¶ 39. The upshot for semiconductor companies is that the first to develop technology for the next smaller chip size-known as a “node,” “process,” or “process node”-gains a significant advantage and can capture a large majority of the revenues for that chip size. Id. ¶ 40. Up through 2011 with the release of its 22nm chip, Intel's chip development had closely followed the progression predicted by Moore's Law, and Intel was a full node ahead of its competitors. Id. ¶ 41.
Although Intel matched the pace of Moore's Law through 2011, it encountered challenges when moving from the 22nm node to the 14nm node. Id. ¶¶ 41-42. Despite planning to begin production of 14nm chips in 2013, Intel was unable to market the chips in large quantities until 2015. Id. ¶ 42. It encountered similar setbacks with the next process node. Id. ¶ 43. Despite the cadence of Moore's Law calling for 10nm chips in 2015, Intel delayed the launch of its 10nm chips to the second half of 2017. Id. Later, it pushed back launch even further into 2019. Id. ¶¶ 43, 45. While Intel dealt with those delays, its competitors began to catch up. By 2018, one of its foundry competitors, TSMC, introduced its 7nm process as Intel continued to work on its 10nm process. Id. ¶ 47. TSMC also formed an alliance with AMD, one of Intel's fabless chip design competitors, allowing AMD to develop increasingly sophisticated chips and seize market share from Intel. Id. ¶¶ 48-51. In an effort to regain ground, Intel hired Jim Keller, a well-regarded microprocessor architect, in April 2018. Id. ¶ 55.
It was against this backdrop of increased competition that Intel began to discuss its upcoming 7nm process with the market. Beginning in May 2019 at an investor meeting, Intel executives, including Swan and Renduchintala, projected that Intel would launch its first 7nm product, known as Ponte Vecchio, in 2021. Id. ¶¶ 60, 62. Renduchintala explained that Intel had learned from the missteps surrounding its 10nm process and that those lessons would allow Intel to meet its anticipated schedule. Id. ¶ 61. Throughout the Class Period, Defendants repeatedly affirmed that Intel's 7nm process was “on track,” reassuring markets that it would meet its 2021 timeline by implementing lessons learned from the 10nm process. Id. ¶¶ 68, 71-72; see also Id. ¶¶ 147, 149, 153, 156, 163, 167, 172, 176, 187.
But, according to Lead Plaintiffs, the 7nm process was not “on track” and had fallen behind schedule while Defendants were making those statements. On December 12, 2019, the technology news website Semi Accurate published an article by Charlie Demerjian reporting that Intel's internal product road maps showed some of its 7nm products were already delayed by at least a year. Id. ¶¶ 75, 77; Decl. of Gina F. Elliott (“Elliott Decl.”), ECF No. 57, Ex. 14. A former Senior Director of Marketing at Intel, FE 1, likewise explained, by December 2019, it was understood at Intel that “yea, 7nm is messed up.” Id. ¶ 78. FE 1 also reported that, before December 2019, Intel's former VP of Marketing told FE 1 the 7nm process was one or two years behind schedule. Id. A former Intel Development Technician and Operations Manager, FE 2, noted that Intel was having yield problems with its 7nm process, meaning that too many of the fabricated chips were defective. Id. ¶ 79. Further, according to another Semi Accurate article by Demerjian, on March 31, 2020, Intel missed a hard tapeout deadline for Ponte Vecchio. Id. ¶ 86. Allegedly, the result of missing the deadline was that Intel would be unable to meet its goal of launching Ponte Vecchio in 2021. Id. ¶¶ 84-87. Moreover, leaks of what appeared to be internal Intel slides from May 2020, which were partially in Russian, indicated that some of Intel's other 7nm products were not scheduled to arrive until 2023. Id. ¶¶ 90-91.
FE 1 disclosed that by May or June 2020, Keller, the microprocessor architect hired by Intel in 2018, had come into conflict with Renduchintala over delays to the 7nm process. Id. ¶ 93. Keller purportedly approached Swan to protest what he viewed as Renduchintala's refusal to address problems with development, and he threatened to resign if Swan did not act. Id. When Swan refused Keller's requests, Keller advised the Intel Board of Directors that neither Swan nor Renduchintala should remain in their roles. Id. ¶¶ 93-94. Shortly thereafter, on June 11, 2020, Keller departed Intel due to what were announced as “personal reasons.” Id. ¶ 95. Analysts reacted with concern, writing that Keller's departure “is a big deal and suggests that whatever he was implementing at Intel was not working or the old Intel guard did not want to implement it,” and interpreting the departure as a sign that “Intel's processor and process node road maps are going to be more in flux or broken than even we had expected.” Id. ¶ 96. Following the news, Intel's stock price fell by 0.6%, declining from a closing price of $59.70 per share on June 11, 2020, to a closing price of $59.33 per share on June 12, 2020. Id. ¶ 99.
On July 23, 2020, Intel issued a press releasing announcing that its 7nm product schedule would be delayed approximately six months due to problems with yield, and Intel's targets for yield were approximately twelve months behind schedule. Id. ¶ 101. In Intel's Q2 2020 earnings call later that day, Swan also disclosed that Intel would utilize outside manufacturers to make some of its leading-edge 7nm chips and that, for some time, it had been working on contingency plans to allow for its chips to be manufactured by third-party foundries if the need arose. Id. ¶¶ 103-04. Such contingency plans required Intel to design its chips to be made in both Intel and non-Intel fabs, an undertaking that would have required eight to twelve months of design work per a former Intel circuit design engineer, FE 3. Id. ¶ 106. Analysts responded negatively to the news, lowering their target prices for Intel and assailing the 7nm delays as a failure. Id. ¶¶ 107-15. Intel's stock price also dropped 17.93% from a closing price of $60.40 per share on July 23, 2020, to a closing price of $49.57 per share on July 27, 2020. Id. ¶ 116.
On October 22, 2020, Intel held its Q3 2020 earnings call in which Swan expanded on his earlier statements about the outsourcing of manufacturing, stating that other 2023 products would also be manufactured both in-house and externally. Id. ¶ 120. After the earnings call, Intel's stock price declined 10.57% from a closing price of $53.90 per share on October 22, 2020, to a closing price of $48.20 per share on October 23, 2020. Id. ¶ 126.
Both Renduchintala and Swan departed Intel following these disclosures-Renduchintala on July 27, 2020, and Swan on February 15, 2021. Id. ¶¶ 117, 128.
II. CHALLENGED STATEMENTS
Lead Plaintiffs challenge the below statements, which are presented chronologically. The Court numbers each statement for ease of reference, and it bolds and italicizes portions of the statements as they were emphasized in the Consolidated Complaint, which presumably was intended to identify the portions of statements that are alleged to be false or misleading:
October 24, 2019 - Q3 2019 Earnings Call
Statement 1: “As we discussed at the May [2019] investor meeting, we are accelerating the pace of process node introductions and moving back to a 2- to 2.5-year cadence. Our process technology and design engineering teams are working closely to ease process design complexity and balance schedule, performance, power and cost. We are on track to launch our first 7-nanometer-based product, a data center-focused discrete GPU, in 2021, 2 years after the launch of 10-nanometer.” Compl. ¶ 132 (alteration in original).
Statement 2: “Last - back in our Analyst Day, we tried to go through this in quite a bit of detail, both, one, kind of our lessons learned coming out of the challenges we had with 10 and how we're capturing those lessons learned as we think about the next 2 generations....” Id. ¶ 134.
Statement 3: “And we indicated that our first product will be 2 years from this quarter, so fourth quarter of2021, our first 7-nanometer product will come out. And our expectation is we'll get back on a 2-year cadence in 7 and beyond. So lots of learnings out of 10-nanometer that we've incorporated. And we said back in May and we reiterated today, we expect to be back to a 2- to 2.5-year cadence going forward, at least for the next few nodes.” Id. ¶ 134.
Statement 4: “Yes. I mean, first to the comment, yes, the - nothing new about process relative to what we said at Analyst Day, ramp 10, 2-year cadence for 7 and our expectations that the cadence going forward will be more like 2- to 2.5-year time frame. So intently focused on 10 now and 7 for the product you mentioned in the fourth quarter. So we're investing to recapture process leadership going forward.” Id. ¶ 136.
Statement 5: “At the same time, we're going to be extremely open-minded about how do we ensure that we're building the best products, and where we build them is something that we'll always evaluate. I think, as you know with the other foundry players, they've been a source of our capacities over the years. And our expectation is, to the extent that they can do something to support our growth better and/or for peak kind of demands, we're always going to look at how do we evaluate the opportunity set that's going to position us best to meet our customers' demand for the growing diversity of products that we have in our portfolio.” Id. ¶ 136.
Statement 6: “ We continue to add capacity in 14-nanometer and began adding capacity at 7-nanometer as well. So we're very focused on getting the capacity in place that will allow us to take the word shortage out of our quarterly discussions.” Id. ¶ 137.
November 4, 2019 - Davis Interview with Barron's
Statement 7: “[W]e're moving to a two to two and a half year cadence on the next nodes. So we're pulling in the spending on 7 nanometer, which will start up in the second half of 2021 ....” Id. ¶ 147.
November 4, 2019 - Benchmark Analyst Report
Statement 8: Benchmark wrote, “[Intel] said that . . . capital efficiency will begin to improve at a faster rate at the 7nm node as EUV is inserted into the manufacturing process. [Intel's] mindset at this point appears to be ‘hit the mark,' meaning that the Company expects to do whatever it takes to meet product schedule expectations. To that end, [Intel] remains committed to node transitions on a 2 to 2.5 year cadence. Interestingly, [Intel] said that it has no interest in being a Foundry any longer as TSMC's business model is very different from [Intel's] and [Intel] expects its IDM model to be intact for the foreseeable future. On the flip-side, [Intel] does not seem very interested in outsourcing capacity needs.” Id. ¶ 149 (alterations in original).
December 3, 2019 - Credit Suisse Conference
Statement 9: “Yes, I mean it's - maybe start with a function of scar tissue. And scar tissue meaning the challenges that we had and the learnings we got from the 22 nanometer to 14 nanometer transition, the 14 nanometer to 10 nanometer transition. In light of that, how are you learning from the past that builds your confidence in the future. * * * So, along the way, we - based on our confidence of past performance, we set a higher and higher bar and it didn't work effectively. Just took too long. And now, good news is we feel like we've got a fairly well dialed in. The bad news is it took too long. . . .” Id. ¶ 153.
Statement 10: “Secondly, we're not going to try to do 2.4 scaling or 2.7 scaling. As we think about 7- nanometer, we put 2.0 back in line with historical trends. As we think about 5-nanometer, which would be our competitor's 3-nanometer, it's more like 2.0 we're thinking about. So, we're not putting as much challenge on the fab and not taking on so much complexity in design rules, which - the more there are the more complicated - the more complicated it is. So we're capturing these learnings from the past and are applying them going forward.” Id. ¶ 153.
Statement 11: “ The third thing, to your point is, with 7-nanometer, one of the benefits of, I should say, of 10-nanometer taking long is we've been playing with EUV for a while. So, this is a new generation of technology. We've been playing with it for a while. While 10-nanometer has took long, our teams including our tool provider had lots of time to work through the inherent challenges of bringing the new technology and list them online.” Id. ¶ 153.
Statement 12: “So, on the combination of learnings from the past, capturing those learnings, how do we think about those going forward, applying them and then tracking along the way. So 7-nanometers we didn't start it yesterday. We started several years ago and we monitor performance on density, on functionality and based on kind of where we are today, we feel pretty good about getting to a two, 2.5 year cadence and launching our first 7-nanometer product in the fourth quarter of 2021.” Id. ¶ 153.
December 10, 2019 - UBS Conference
Statement 13: “But first of all, as somebody who regards themselves of the technologies first and foremost, you go through your career very much understanding that, your most seminal learnings come from the programs or the activities you were part of that didn't go according to plan. And with 10 nanometers, I think the company has learned a number of really, really crucial lessons that I think sets us up to be a much, much more mature decision-making organization going forward.” Id. ¶ 156.
Statement 14: “ I would say, on 10, we learned four key lessons. The first is really to balance the pursuit of scaling and cost together with schedule predictability power and performance. And Intel was very focused on continuing to achieve a cost per transistor curve, that complied with the Moore's Law cadence of every 2 years. * * * So going forward on 7, we've taken a much more pragmatic approach of basically making sure the scaling risk we take doesn't perturbate our ability to deliver to schedule and to power and performance. So as we've said in the past, we've moved more to a scaling factor around 2x between 10 and 7.” Id. ¶ 156.
Statement 15: “ The other thing that we learned and this was very much a lesson we learned on 14 is to continue to harvest intra-node product optimizations or intra-node process optimizations....” Id. ¶ 156.
Statement 16: “The third [lesson learned] is to maintain a mix of nodes, going forward as well not out of one node into another node in full gusto, but essentially take a much more gradual and measured migration between nodes, because not every part of technology that we deliver in an SoC benefits proportionately from logic scaling. So, for concerns of time to market, power and performance and margin maintaining a mix of nodes going forward in a heterogeneous product construction approach is really important.” Id. ¶ 156 (alteration in original).
Statement 17: “And the final lesson, probably one of the most important lessons is to make it easy andfast for our design teams to be able to migrate through intra-node transitions. For us one of the key things we've really done is to make sure that as we've delivered process goodness, we've also made that much easier for our design teams to pick up and run with so that we can get much greater velocity in our product cadence. So all of those have been integrated into our approach on 7. And I think we're making good progress on 7 as a result of that. And as we've announced previously, we'll see our
first 7 nanometer product shipping in 2021 with a full portfolio in 2022.” Id. ¶ 156.
Statement 18: “[T]he GPGPU, or GPU in general benefits from the scaling and performance and power advantage that come with the transition from 10 to 7 as significantly as for example as CPU. And thirdly, [a] GPGPU by nature of its architecture and redundancy in the architecture makes it a lot more amenable to being a good ramp vehicle in the early phases of a new node where defect density is still being pushed down to its absolute minimum. But there won't be a large gap between the launch of our first product in 7 and the rest of our portfolio. So you can expect a full portfolio of products across our entire product portfolio within a year of that first product launch.” Id. ¶ 156 (alterations in original).
Statement 19: “Yeah, well, first of all, I think we regard companies like TSMC and Samsung as strategic partners. Intel's had a long history, over two decades of history of working with the foundry ecosystem. And in fact, something like 20% to 25% of the wafer volume that we source comes from outside of the Company and we don't see that changing in any major fashion going forward.” Id. ¶ 157.
Statement 20: “But that said, we still believe that there is tremendous value in the IDM approach we have going forward. And if you look at the assets that Intel brings to bear, we have process technology, we have advanced packaging technology; we have memory technology; we have interconnect technology, we also have an incredibly important franchises at CPU, which is a cornerstone IP; we're building a portfolio of what we call xPU architectures like the GPU, the FPGA, the neural network processes. And we're integrating that with really strong focus on both security and harmonizing software. And if you integrate all of that together, you get an incredibly potent innovation environment, that's very difficult to replicate in a fabless foundry partnership. So while we think, there is great value engaging with and learning from the external foundry ecosystem, we still think that there is tremendous generate - a value we can generate by continuing to be an IDM. So we play those positions intelligently and pragmatically to deliver the best portfolio we can for our customers.” Id. ¶ 157.
January 23, 2020 - Q4 2019 Earnings Call
Statement 21: “Our 7-nanometer process remains on track to deliver our lead 7-nanometer product, Ponte Vecchio, at the end of2021, with CPU products following shortly after in 2022.” Id. ¶ 163.
January 24, 2020 - 2019 Form 10-K3
Statement 22: “ We are accelerating the pace of process node introductions and moving back to a 2- to 2.5-year cadence. We are on track to deliver our first 7nm-based product, a discrete GPU, at the end of 2021.” Id. ¶ 167.
Statement 23: “ We are an IDM. Unlike many other semiconductor companies, we primarily design and manufacture our products in our own manufacturing facilities, and we see our in-house manufacturing as an important advantage. We continue to develop new generations of manufacturing process technology as we seek to realize the benefits from Moore's Law. Realizing Moore's Law results in economic benefits as we are able to either reduce a chip's cost as we shrink its size, or increase functionality and performance of a chip while maintaining the same cost with higher density. This makes possible the innovation of new products with higher performance while balancing power efficiency, cost, and size to meet customers' needs. Our ability to optimize and apply our manufacturing expertise to deliver more advanced, differentiated products is foundational to our current and future success.” Id. ¶ 167.
Statement 24: “ We are on track to deliver our first 7nm-based product, a data center-focused discrete GPU, at the end of2021. We are approaching next-generation process nodes with a focus on striking an optimal balance between schedule, performance, power, and cost and will continue to drive intra-node advancement.” Id. ¶ 167.
Lead Plaintiffs also allege that Swan and Davis's Sarbanes-Oxley (“SOX”) certification was false and misleading because Intel's Form 10-K contained false and misleading statements. Compl. ¶¶ 168, 171. Because Lead Plaintiffs' theory of falsity for the SOX certification is predicated on the falsity of other challenged statements in Intel's Form 10-K, the certification will rise or fall with those other statements, and the Court does not address it separately.
March 2, 2020 - Morgan Stanley Conference
Statement 25: “But so I feel like we're in the 10-nanometer node. It's important that we're continuing to see yield improvements ratably over the time period. But as we said back in our Analyst Day in May of ‘19, look this isn't going to be the best node that Intel has ever had. It's going to be less productive than 14, less productive than 22, but we're excited about the improvements that we're seeing and we expect to start the 7-nanometer period . . . with a much better profile of performance over that starting at the end of ‘21.” Id. ¶ 172.
Statement 26: “ Yeah, I think we feel very good about where the road map is going. . . . we feel like we're starting to see the acceleration on the process side that we have been talking about to get back to parity in the 7-nanometer generation and regain leadership in the front down there.” Id. ¶ 172.
April 23, 2020 - Q1 2020 Earnings Call
Statement 27: “On the second part of your question, I'd go back to the commentary that George provided back at our Analyst Day in the spring, which is, obviously, when we transition from a mature node to a new node, margins tend to come down. We indicated that we plan to get back on a 2- to 2.5-year cadence, which means in 2021, we'll be ramping 10-nanometer even more while we're investing in 7-nanometer that we anticipate having in the fourth quarter of 2021. So those dynamics of - from a mature node to a new node, impacts the gross margins of the business, but we feel like it's - we're well on track from the plans we laid out and feel pretty good about a dynamite first quarter and an outlook for the second quarter in line or better than what we expected.” Id. ¶ 176.
June 11, 2020 - Press Release Regarding Keller's Departure
Statement 28: “Today, Intel announced that Jim Keller has resigned effective June 11, 2020, due to personal reasons. Intel appreciates Mr. Keller's work over the past two years helping them continue advancing Intel's product leadership and they wish him and his family all the best for the future. Intel is pleased to announce, however, that Mr. Keller has agreed to serve as a consultant for six months to assist with the transition.” Id. ¶ 183.
June 11, 2020 - Deutsche Bank Analyst Report
Statement 29: Deutsche Bank reported that Swan and Davis stated they were “[l]ooking forward to 7nm, [Intel]'s time-line remains unchanged with a late 2021 launch.” Id. ¶ 184 (alterations in original).
June 25, 2020 - Statement to Consumer Electronics Daily
Statement 30: Consumer Electronics Daily published an article stating, “[a]n Intel spokesperson emailed Wednesday [i.e., June 24, 2020] that its 7-nanometer process ‘remains on track' with first products due by the end of 2021.” Id. ¶ 187 (alterations in original).
These statements can be sorted into four general categories, though some fall under more than one category:
1. Statements regarding Intel's 7nm development time line (Statements 1, 3, 4, 7, 8, 12, 17, 18, 21, 22, 24-27, 29, 30);
2. Statements regarding how Intel incorporated “lessons learned” from its 10nm process into development of its 7nm process (Statements 2, 3, 9-17);
3. Statements that Intel remained an IDM (Statements 4-6, 8, 19, 20, 23); and
4. The statement about Keller's departure (Statement 28).
REQUEST FOR JUDICIAL NOTICE
I. LEGAL STANDARD
Ordinarily, a court may not examine materials outside the pleadings when considering a motion to dismiss for failure to state a claim. Lee v. City of L.A., 250 F.3d 668, 688 (9th Cir. 2001). There are two exceptions to this general rule. First, courts may take judicial notice of certain facts that are “not subject to reasonable dispute” because they are “generally known” or “can be accurately and readily determined from sources whose accuracy cannot reasonably be questioned.” Fed.R.Evid. 201(b). Second, the doctrine of incorporation by reference permits courts to treat a document as if it were “part of the complaint itself,” but only if the complaint “refers extensively to the document or the document forms the basis of the plaintiff's claim.” Khoja v. Orexigen Therapeutics, Inc., 899 F.3d 988, 1002 (9th Cir. 2018) (citation omitted).
The Ninth Circuit has cautioned against the use of judicial notice or incorporation by reference to raise expansive factual disputes at the pleading stage. See Khoja, 899 F.3d at 998-99, 1003. A court may take judicial notice of the existence and contents of a public record but may not take notice of the truth of any disputed facts within that record. Id. at 999-1000. Likewise, a court may generally “assume [an incorporated document's] contents are true for purposes of a motion to dismiss under Rule 12(b)(6),” id. at 1003 (alteration in original) (quoting Marder v. Lopez, 450 F.3d 445, 448 (9th Cir. 2006)), but should not assume the truth of facts in an incorporated document “if such assumptions only serve to dispute facts stated in a well-pleaded complaint.” Id.
II. DISCUSSION
Defendants request judicial notice and incorporation by reference of Exhibits 1 through 26 to the Declaration of Gina F. Elliott. Req. for Judicial Notice (“RJN”), ECF No. 59. Lead Plaintiffs do not object to the Court's consideration of Exhibits 1, 7-9, 11-13, 19, 20, and 24-26. Resp. to RJN, ECF No. 65, at 1. However, Lead Plaintiffs object to Exhibits 2-6, 10, 14-18, and 21-23, acknowledging that the Court may take notice of the existence and contents of those exhibits but challenging Defendants' use of those exhibits as improper. Id.
Though Lead Plaintiffs indicate that they object to Exhibit 19, their RJN briefing contains no argument about that exhibit. Consequently, the Court finds that any objection to Exhibit 19 is waived, and in any case, the exhibit is subject to incorporation by reference because it contains one of the challenged statements. See Khoja, 899 F.3d at 1002.
Exhibits 2-6 are certain of Intel's SEC filings, which courts routinely take notice of in federal securities actions. See Metzler Inv. GMBH v. Corinthian Colls., Inc., 540 F.3d 1049, 1064 n.7 (9th Cir. 2008); Kipling v. Flex Ltd., No. 18-CV-02706-LHK, 2020 WL 2793463, at *7 (N.D. Cal. May 29, 2020). Lead Plaintiffs argue, though, that the Court should decline to take judicial notice because the exhibits are irrelevant and do not contain any challenged statement. Resp. to RJN at 3-4. Their argument is unavailing. The PSLRA extends safe harbor protections to oral forward-looking statements that identify cautionary language in readily available written documents. 15 U.S.C. § 78u-5(c)(2)(B). Each of Exhibits 2-6 contains cautionary language referenced by oral statements which Lead Plaintiffs challenge in this action, so the exhibits are relevant to the Court's analysis of whether safe harbor protections apply. See, e.g., Elliott Decl., Ex. 7 at 2 (referring to Exhibit 4, which in turn refers to Exhibit 2). The Court will therefore consider Exhibits 2-6 but will not take notice of any disputed facts.
Exhibits 14-16 are articles from Semi Accurate that Lead Plaintiffs cite in their complaint to establish the falsity of several challenged statements. See Compl. ¶¶ 77, 86, 91. These exhibits are both judicially noticeable as publicly available articles, Heliotrope Gen., Inc. v. Ford Motor Co., 189 F.3d 971, 981 n.18 (9th Cir. 1999), and incorporated by reference as the basis for Lead Plaintiffs' allegations of falsity. See Khoja, 899 F.3d at 1002. Lead Plaintiffs object to Defendants' use of the exhibits, arguing that Defendants are improperly citing the exhibits to argue Semi Accurate is an unreliable source. Resp. to RJN at 7-8. As the Court discusses in more detail below, it is appropriate for a court to assess the reliability of news articles under the PSLRA's heightened pleading standard, and the context of the full articles is relevant to that assessment. See In re McKesson HBOC, Inc. Sec. Litig., 126 F.Supp.2d 1248, 1272 (N.D. Cal. 2000) (requiring a newspaper article to contain “numerous factual particulars” and indications of “an independent investigative effort” before crediting the article for purposes of a scienter analysis). Thus, the Court considers Exhibits 14-16 for the purpose of determining the reliability of the Semi Accurate articles, but it does not assume the truth of facts within those exhibits.
Exhibit 17, which contains pages from the Semi Accurate website that Lead Plaintiffs do not cite in their complaint, does not serve the same purpose. While considering Exhibits 14-16 would provide relevant context as to their reliability, and incorporation by reference of those articles helps to avoid selective citation, see Khoja, 899 F.3d at 1002, the uncited webpages in Exhibit 17 are too far removed from the cited articles to function as context. Rather, Exhibit 17 creates a factual dispute over Lead Plaintiffs' allegations of reliability, see Compl. ¶¶ 75-76, and therefore the Court will not consider Exhibit 17.
Exhibits 10, 18, and 21-23 are the remaining exhibits which Lead Plaintiffs object to. They consist of a conference transcript, online article, and analyst reports, all of which are subject to judicial notice. See In re Facebook, Inc. Sec. Litig., 477 F.Supp.3d 980, 1009 (N.D. Cal. 2020) (transcripts); Heliotrope, 189 F.3d at 981 n.18 (articles); Costanzo v. DXC Tech. Co., No. 19-cv-05794-BLF, 2020 WL 4284838, at *3-4 (N.D. Cal. July 27, 2020) (analyst reports). Exhibits 10 and 18 are also subject to incorporation by reference as sources of Lead Plaintiffs' challenged statements or allegations of falsity. See Compl. ¶¶ 91-92, 153; Khoja, 899 F.3d at 1002. Once again, though, Lead Plaintiffs object to how Defendants use these exhibits in their motion. Resp. to RJN at 4-5, 8-12. Such objections only limit how the Court may use these documents; they do not restrict the Court's ability to take notice of or incorporate the documents. See Bos. Ret. Sys. v. Uber Techs., Inc., No. 19-cv-06361-RS, 2020 WL 4569846, at *3 (N.D. Cal. Aug. 7, 2020). The Court will consider Exhibits 10, 18, and 21-23 for the purpose of showing what information was available to the stock market, but not for the truth of any fact asserted.
Accordingly, the Court takes notice of and/or incorporates by reference all exhibits attached to the Elliott Declaration except for Exhibit 17.
MOTION TO DISMISS
I. LEGAL STANDARD
To survive a motion to dismiss under Rule 12(b)(6), a plaintiff must plead each claim with enough specificity to “give the defendant fair notice of what the . . . claim is and the grounds upon which it rests.” Bell Atl. Corp. v. Twombly, 550 U.S. 544, 555 (2007) (citation omitted). A bare recital of the elements of a claim, supported only with conclusory allegations, is inadequate. Ashcroft v. Iqbal, 556 U.S. 662, 678 (2009). Rather, the complaint must contain sufficient factual allegations to allow a court to reasonably infer that the defendant is liable. Id.
Securities fraud cases must also meet the higher bar set by the particularity requirements of Rule 9(b) and the PSLRA. Or. Pub. Emps. Ret. Fund v. Apollo Grp. Inc., 774 F.3d 598, 604 (9th Cir. 2014). Rule 9(b) requires a plaintiff alleging fraud to plead with particularity the circumstances constituting fraud. Fed.R.Civ.P. 9(b). Specifically, a plaintiff must plead the “who, what, when, where, and how” of the alleged fraud. Kearns v. Ford Motor Co., 567 F.3d 1120, 1124 (9th Cir. 2009) (citation omitted). The PSLRA demands even more, requiring a plaintiff “to state with particularity both the facts constituting the alleged violation and the facts evidencing scienter.” In re Rigel Pharms., Inc. Sec. Litig., 697 F.3d 869, 876 (9th Cir. 2012). To plead falsity, a securities plaintiff must “specify each statement alleged to have been misleading [and] the reason or reasons why the statement is misleading.” Id. at 877 (quoting 15 U.S.C. § 78u-4(b)(1)). To plead scienter, the plaintiff must “state with particularity facts giving rise to a strong inference that the defendant acted with the required state of mind.” Id. (quoting 15 U.S.C. § 78u-4(b)(2)(A)). An inference of scienter must be more than plausible, it must be “cogent and at least as compelling as any opposing inference of non fraudulent intent.” Tellabs, Inc. v. Makor Issues & Rights, Ltd., 551 U.S. 308, 314 (2007).
At the pleading stage, the Court accepts all factual allegations as true and construes the pleadings in the light most favorable to the plaintiff. Reese, 643 F.3d at 690. The Court is not required to accept as true “allegations that are merely conclusory, unwarranted deductions of fact, or unreasonable inferences.” In re Gilead Scis. Sec. Litig., 536 F.3d 1049, 1055 (9th Cir. 2008) (quoting Sprewell v. Golden State Warriors, 266 F.3d 979, 988 (9th Cir. 2001), amended on other grounds, 275 F.3d 1187 (9th Cir. 2001)).
II. SECTION 10(B) AND RULE 10B-5
To state a claim under Section 10(b) and Rule 10b-5, a plaintiff must allege “(1) a material misrepresentation or omission by the defendant; (2) scienter; (3) a connection between the misrepresentation or omission and the purchase or sale of a security; (4) reliance upon the misrepresentation or omission; (5) economic loss; and (6) loss causation.” In re Quality Sys., Inc. Sec. Litig., 865 F.3d 1130, 1140 (9th Cir. 2017) (quoting Halliburton Co. v. Erica P. John Fund, Inc., 134 S.Ct. 2398, 2407 (2014)). Forward-looking statements “may still be shielded from liability by the ‘safe harbor' provision of the PSLRA” even if all elements of a Section 10(b) claim are pleaded. Quality Sys., 865 F.3d at 1141; 15 U.S.C. § 78u-5(c).
In their motion to dismiss, Defendants argue that many of the challenged statements fall under the PSLRA safe harbor, and that Lead Plaintiffs have failed to plead actionable misstatements or omissions, a strong inference of scienter, and loss causation. Mot. at 1.
A. PSLRA Safe Harbor
Under the PSLRA safe harbor, a forward-looking statement is not actionable under federal securities law if it is (a) identified as forward-looking and accompanied by meaningful cautionary language or (b) was made without actual knowledge that it was false or misleading. 15 U.S.C. § 78u-5(c)(1). The safe harbor operates disjunctively, so a forward-looking statement is protected if either condition is met. Wochos v. Tesla, Inc., 985 F.3d 1180, 1190 (9th Cir. 2021) (citing Quality Sys., 865 F.3d at 1141; In re Cutera Sec. Litig., 610 F.3d 1103, 1112-13 (9th Cir. 2010)). For oral forward-looking statements, the safe harbor applies if the speaker or someone acting on her behalf warns that actual results may differ, and she directs the audience to cautionary language in a readily available written document. 15 U.S.C. § 78u-5(c)(2).
A statement is forward-looking if it is about “(1) financial projections, (2) plans and objectives of management for future operations, (3) future economic performance, or (4) the assumptions ‘underlying or related to' any of these issues.” No. 84 Emp.-Teamster Joint Council Pension Tr. Fund v. Am. W. Holding Corp., 320 F.3d 920, 936 (9th Cir. 2003) (citing 15 U.S.C. § 78u-5(i)). A forward-looking statement may be mixed with statements of present or past fact.
In that case, the forward-looking aspects of the mixed statement are protected by the safe harbor, but portions addressing the present or past are not. Quality Sys., 865 F.3d at 1142. The non forward-looking aspects of mixed statements are actionable only if they are “separable” from the forward-looking aspects and contain a “concrete assertion concerning a specific current or past fact” that “goes beyond the articulation of plans, objectives, and assumptions.” Wochos, 985 F.3d at 1190-91 (cleaned up) (citations omitted)
1. Statements Regarding Intel's 7nm Development Timeline
a. Forward-Looking Statements
Defendants argue that all statements regarding Intel's 7nm timeline are forward-looking. Mot. at 10-12. The Court agrees. The statements which set forth Intel's projected launch date for its 7nm products, or which indicate Intel's expectations of a two-year development cadence between process nodes, are plainly forward-looking statements of plans and objectives. See 15 U.S.C. § 78u-5(i)(1)(B); Am. W. Holding Corp., 320 F.3d at 936. The remaining statements asserting that Intel was “on track” to meet those goals are likewise forward-looking under Wochos. 985 F.3d at 1192.
Lead Plaintiffs largely do not contest that statements about Intel's 7nm timeline are forward-looking, identifying only two statements that they consider to be about present fact: Statement 26 (“we feel very good about where the [7nm] road map is going”) and Statement 29 (“[the 7nm] time-line remains unchanged”). Opp'n, ECF No. 64, at 12-13. In their view, these are remarks about the present state of Intel's road map and time line rather than about Intel's future objectives. Id. However, Wochos considered and rejected this exact argument. The plaintiffs in Wochos argued that “on track” statements were not forward-looking because they concerned the present state of progress towards a goal. 985 F.3d at 1191. The court disagreed, finding the “on track” statements to be forward-looking on the grounds that they simply reasserted previously announced future objectives. Id. at 1192. That is so, the court explained, since an announced goal “necessarily reflects an implicit assertion that the goal is achievable based on current circumstances.” Id. As a result, “an unadorned statement” that a company will be able to achieve its objective is “merely [an] alternative way[] of declaring or reaffirming the objective itself.” Id. Statements 26 and 29 are such “unadorned statements” because they do no more than convey that Intel's 7nm launch goals are achievable under current circumstances, and they are therefore forward-looking.
Lead Plaintiffs also briefly argue that Statements 26 and 29 are mixed statements, but they do not identify any purported assertions of current or past fact other than those about Intel's road map and timeline. Opp'n at 13.
b. Meaningful Cautionary Language
i. Written Statements
As statements about Intel's 7nm timeline are forward-looking, they are protected under the safe harbor if they are accompanied by meaningful cautionary language. The Court begins with Statements 22 and 24, which are written statements in Intel's 2019 Form 10-K. To be meaningful, cautionary language 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). These factors must be “substantive company-specific warnings based on a realistic description of the risks applicable to the particular circumstances.” Rodriguez v. Gigamon Inc., 325 F.Supp.3d 1041, 1052 (N.D. Cal. 2018) (quoting In re Harman Int'l Indus., Inc. Sec. Litig., 791 F.3d 90, 102 (D.C. Cir. 2015)). But the cautionary language “does not need to warn of the ‘exact risk' that transpires.” Bodri v. GoPro, Inc., 252 F.Supp.3d 912, 931 (N.D. Cal. 2017).
Intel included extensive risk factors with its 2019 Form 10-K. Elliott Decl., Ex. 1 at 50-60. As pertinent to the statements about Intel's 7nm timeline, Intel warned that product delays could occur and had occurred, that such delays could harm company performance, and that yields might be low:
[T]o the extent we do not timely introduce new manufacturing process technologies that improve transistor density with sufficient manufacturing yields and operational efficiency, relative to competing foundry processes, we can face cost and product performance disadvantages.
* * *
We are not always successful or efficient in developing or implementing new process nodes and production processes. For
example, we experienced significant delays in implementing our 10nm process technology....
Risks inherent in the development of next-generation process technologies include production timing delays [and] lower-than-anticipated manufacturing yields .... Production timing delays have at times caused us to miss customer product design windows, which can result in lost revenue opportunities and damage to our customer relationships. Furthermore, when the introduction of next-generation process nodes is delayed, including additional competitive features in our products can result in larger die size products, manufacturing supply constraints, and increased product costs. Lower manufacturing yields and longer manufacturing throughput times, compared to previous process nodes, can increase our product costs and adversely affect our gross margins .... In addition, as the die size of our products has increased and our manufacturing process nodes have shrunk, our products and manufacturing processes have grown increasingly complex and more susceptible to product defects and errata, which can also contribute to production timing delays and lower yields.
From time to time, disruptions in the production process result from errors . . . which could affect the timing of production ramps and yields.... [T]o the extent we face delays in the timing of our product introductions, we could become less competitive and lose revenue opportunities, and our gross margin could be adversely affected ....Id. at 51, 54-55. This cautionary language warns of the exact event that Lead Plaintiffs allege Defendants concealed (the delay of Intel's 7nm process) as well as the underlying cause and “primary driver” of that event (poor yields). See Compl. ¶ 101. That is more than enough to be meaningful. See In re Pivotal Sec. Litig., No. 3:19-cv-03589-CRB, 2020 WL 4193384, at *16 (N.D. Cal. July 21, 2020) (cautionary language that “addressed the very subjects Plaintiffs challenge” was meaningful); Kipling, 2020 WL 2793463, at *12-13 (cautionary language was meaningful when it discussed “operational difficulties [] of the same nature as the ones that Plaintiff alleges rendered the forward-looking statements false”).
Lead Plaintiffs raise three arguments for why the cautionary language in Intel's 2019 Form 10-K is not meaningful, but none of their arguments is availing. Their first two arguments-that the risk factors are boilerplate, and that the risk factors arranged under the heading, “Changes in Product Demand Can Adversely Affect Our Financial Results: We face significant competition,” are irrelevant-may be quickly disposed of. Opp'n at 15. The risk factors identified above warn of Lead Plaintiffs' exact theory of falsity, so they are far from boilerplate. And the method by which Intel organizes and labels its risk factors has no impact on whether the cautionary language is meaningful. The risk factors above are clearly relevant, and the heading criticized by Lead Plaintiffs does not otherwise cause any confusion about their meaning.
The third argument requires more attention. Lead Plaintiffs argue that the cautionary language cannot be meaningful because it “presented the risks as merely possible when they had already materialized.” Id. at 15-16 (emphasis removed). While they are correct that cautionary language may not be meaningful if it suggests that risks have not been realized when they have already occurred, Washtenaw Cnty. Emps. Ret. Sys. v. Celera Corp., No. 5:10-cv-02604-EJD, 2012 WL 3835078, at *4 (N.D. Cal. Sept. 4, 2012), that is not what happened here. Intel's risk factors make clear that product delays, the very risk that Lead Plaintiffs allege to have materialized, occurred in the past. Intel flagged that “[p]roduction timing delays have at times caused us to miss customer product design windows.” Elliott Decl., Ex. 1 at 54. By disclosing that delays “have at times” caused issues, Intel indicated that such delays had occurred before. What is more, Intel identified a specific instance of product delays materially impacting its ability to compete, noting that “we experienced significant delays in implementing our 10nm process technology” and that “our delays in transitioning to this node occurred while third-party foundries developed new, competitive process technologies.... which can help increase the competitiveness of their products.” Id. In light of these disclosures, the Court concludes that Statements 22 and 24 are accompanied by meaningful cautionary language and so protected by the PSLRA safe harbor.
ii. Oral Statements
The Court next turns to Defendants' oral statements. Unlike for written statements, the cautionary language accompanying oral statements does not need to identify specific risk factors. 15 U.S.C. § 78u-5(c)(2). Instead, forward-looking oral statements receive safe harbor protection if they are accompanied by a general statement that results might differ, and an additional statement directing the audience to a readily available written document with more detailed risk factors. Id. The detailed factors in that written document, in turn, must constitute meaningful cautionary language. 15 U.S.C. § 78u-5(c)(2)(B)(iii).
Statements 1, 3, 4, 12, 21, and 25-27 fall under this part of the PSRLA safe harbor. Specifically, Statements 1, 3, 4, 21, and 27 were made on earnings calls that began with the following: “Before we begin, let me remind everyone that today's discussion contains forward looking statements based on the environment as we currently see it and as such does include risks and uncertainties. Please refer to our press release for more information on the specific risk factors that could cause actual results to differ materially.” Elliott Decl., Ex. 7 at 2 (Statements 1, 3, 4); see also Elliott Decl., Ex. 8 at 3 (Statement 21); Ex. 9 at 3 (Statement 27). Statements 12, 25, and 26 were made at investor conferences which opened with similar language: “Today's presentation may contain forward-looking statements. All statements . . . that are not historical facts are subject to a number of risks and uncertainties, and actual results may differ materially. Please refer to their more recent earnings release, Form 10-Q and Form 10-K for more information on the specific risk factors that could cause actual results to differ.” Elliott Decl., Ex. 10 at 1 (Statement 12); see also Elliott Decl., Ex. 12 at 1 (Statements 25 and 26).
This cautionary language accompanying Defendants' oral statements satisfies the requirements of the PSLRA safe harbor. The Ninth Circuit and district courts in the circuit have repeatedly approved of cautionary language similar to that preceding Defendants' statements on earnings calls and at investor conferences. See, e.g., Police Ret. Sys. of St. Louis v. Intuitive Surgical, Inc., 759 F.3d 1051, 1059 (9th Cir. 2014) (“Before we begin, I would like to inform you that comments mentioned on today's call may be deemed to contain forward-looking statements. Actual results may differ materially from those expressed or implied, as a result of certain risks and uncertainties. These risks and uncertainties are described in detail in the company's [SEC] filings. Prospective investors are cautioned not to place undue reliance on such forward-looking statements.” (alteration in original)); Kipling, 2020 WL 2793463, at *11 (“This meeting and these presentations contain forward-looking statements, which are based on current expectations and assumptions that are subject to risks and uncertainties, and actual results could materially differ. Such information is subject to change and we undertake no obligation to update these forward-looking statements. For a discussion of the risks and uncertainties, see our most recent filings with the SEC, including our current annual and quarterly reports.”); Barry v. Colony North Star, Inc., No. CV 18-2888-GW (MRWx), 2022 WL 17084923, at *13 (C.D. Cal. Feb. 10, 2022) (collecting cases). The Court therefore finds that Intel's cautionary language is sufficient.
The Court also concludes that the documents to which Intel's warnings cite-Intel's Forms 10-K and earnings press releases-contain meaningful cautionary language, as required. The warnings accompanying Statements 25 and 26 cite to Intel's 2019 Form 10-K, see Elliott Decl., Ex. 12 at 1, which the Court has already found to contain meaningful cautionary language. The warnings accompanying Statement 12 cite to an earlier SEC filing, Intel's 2018 Form 10-K. Elliott Decl., Ex. 10 at 1. Although the cautionary language in the 2018 Form 10-K differs somewhat from that of the 2019 Form 10-K, it still warns of both production delays and low yields, and it highlights Intel's 10nm delays as an example of a delay. Elliott Decl., Ex. 2 at 52. As such, that language is still meaningful.
Unlike the two warnings made during investor conferences, the warnings that accompanied Statements 1, 3, 4, 21, and 27 on earnings calls did not directly reference SEC filings. Rather, they referred to Intel's earnings press releases. Elliott Decl., Ex. 7 at 2; Ex. 8 at 3; Ex. 9 at 3. The press releases contained abbreviated cautionary language advising of risks related to “the timing of qualifying products for sale,” “the timing of Intel product introductions,” and “variations related to . . . product manufacturing quality/yields.” Elliott Decl., Ex. 4 at 4; see also Elliott Decl., Ex. 5 at 4; Ex. 6 at 4. These risk factors inform investors of the possibility of product delays and yield issues, though they do not specify that Intel had previously experienced delays. Nonetheless, the Court finds that the press releases contain meaningful cautionary language because they direct investors to Intel's most recent Forms 10-K, which do reveal that Intel had previously dealt with product delays. Elliott Decl., Ex. 4 at 5; Ex. 5 at 5; Ex. 6 at 5.
Disputing that the cautionary language is adequate, Lead Plaintiffs argue that oral statements are not protected by references to cautionary language that is “scattered” between multiple SEC filings, and that Intel's 2018 Form 10-K failed to warn specifically about Intel's 7nm process. Opp'n at 13-14.
In support of their first argument, Lead Plaintiffs cite to In re HI/FN, Inc. Securities Litigation, which held that “misleading oral statements are not protected by cautionary language spread out among various documents.” No. C-99-4531 SI, 2000 WL 33775286, at *5 (N.D. Cal. Aug. 9, 2000) (cleaned up). In re HI/FN is in apposite because it applied the judicially created bespeaks caution doctrine rather than the statutory PSLRA safe harbor. Id.; see also Barry, 2022 WL 17084923, at *12 (distinguishing between the bespeaks caution doctrine and PSLRA safe harbor). The safe harbor expressly permits defendants to invoke its protections for oral statements by cross-referencing cautionary language in written statements. Intel did that here, thereby satisfying the requirements of the safe harbor. Lead Plaintiffs' second argument is also to no effect. The PSLRA safe harbor demands only that companies warn of risks that might cause actual results to differ from forward-looking predictions; nothing in the statutory language obliges a company to specifically name its products. See 15 U.S.C. § 78u-5(c)(1)(A); Gammel v. Hewlett-Packard Co., 905 F.Supp.2d 1052, 1067 (C.D. Cal. 2012) (finding cautionary language to be meaningful even when it did not expressly refer to the product at issue).
Lead Plaintiffs also cite Tarapara v. K12 Inc., No. 16-cv-4069-PJH, 2017 WL 3727112 (N.D. Cal. Aug. 30, 2017), arguing that cautionary statements made at different times than the challenged statements do not trigger safe harbor protections. Opp'n at 14 (citing Tarapara, 2017 WL 3727112, at *13). In that case, though, there was no indication that the oral statements were accompanied by an explicit cross-reference to the written documents containing detailed risk factors, as required by the PSLRA. Since such cross-references are present here, Tarapara is inapposite.
c. Actual Knowledge
Statements 7, 8, 17, 18, and 30 are not accompanied by any cautionary language, so they are protected by the PSLRA safe harbor only if Defendants did not have actual knowledge that those statements were false. 15 U.S.C. § 78u-5(c)(1)(B). Defendants argue that, to plead actual knowledge, Lead Plaintiffs needed to show that Defendants knew it was impossible to achieve their goals for Intel's 7nm timeline. Mot. at 15 (citing Wochos, 985 F.3d at 1194). The analysis in Wochos is instructive for how this Court should assess scienter as to forward-looking goals, but Wochos did not require knowledge of impossibility. In that case, the Ninth Circuit was not addressing actual knowledge when it considered whether defendants knew it was impossible to reach their goals; instead, it dealt with the question of whether knowledge of impossibility would render cautionary language not meaningful. Wochos, 985 F.3d at 1193-94. And even then, the Ninth Circuit reserved that question, finding that plaintiffs failed to plead impossibility. Id.
Statements 13-20 are from the December 10, 2019 UBS Conference. Elliott Decl., Ex. 11. While the transcript of the event contains cautionary language, it appears that the language was added after the fact by the transcription service. Id. at 10. It does not seem that any cautionary language was given during the conference itself, so the Court cannot conclude that the statements were accompanied by cautionary language.
Nonetheless, the Court concludes that Lead Plaintiffs have not pleaded actual knowledge for the reasons given in its discussion of scienter below. Accordingly, Statements 7, 8, 17, 18, and 30-and all other forward-looking statements regarding Intel's 7nm timeline-are protected by the safe harbor.
* * *
In conclusion, Statements 1, 3, 4, 7, 8, 12, 17, 18, 21, 22, 24-27, 29, and 30 are each protected by the PSRLA safe harbor to the extent they are about Intel's 7nm timeline or product development cadence.
2. Statements Regarding Lessons Learned
a. Forward-Looking Statements
Defendants argue that statements about lessons learned are forward-looking under Wochos. Mot. at 12. From their perspective, a statement about lessons learned is an assumption about how the progression of events leading up to a future goal will play out. Reply, ECF No. 67, at 2. Lead Plaintiffs counter that Wochos did not hold that “lessons learned” statements are categorically forward-looking, and that statements about lessons learned cannot be forward-looking unless they are made in response to questions eliciting forward-looking information. Opp'n at 9-10.
On this point, Defendants overstate Wochos. Unlike an “on track” statement, which is forward-looking because it restates or reaffirms a future goal, a “lessons learned” statement does not necessarily possess the “sort of features that are inherent in any forward-looking statement,” e.g., it is not “an implicit assertion that the goal is achievable based on current circumstances.” Wochos, 985 F.3d at 1192. The Court cannot conclude that all “lessons learned” statements are inherently forward-looking like “on track” statements, and similarly, the Court cannot conclude that all statements about lessons learned fall under the umbrella of safe harbor as assumptions about future events. Wochos distinguished between assumptions about future events, which are protected by safe harbor, and assumptions based on present or past facts, which are not protected. For example, the Ninth Circuit recognized that a company “can readily announce an objective without stating, for example, that the reason why it is achievable is because production of relevant units actually rose 75% over the last quarter or because the company has actually hit certain intermediate benchmarks.” Id. Such factual assertions are outside the safe harbor and can be actionable if they are false. Id. A statement about lessons learned can easily land outside the safe harbor since there is nothing inherently forward-looking about lessons learned from past experiences, and a company can announce its objectives without justifying them with lessons learned. In this context, Wochos stands only for the proposition that statements about lessons learned can be forward-looking, but whether that is so depends on the context.
The Court therefore commences its analysis by examining the “lessons learned” statement in Wochos. There, plaintiffs challenged a response by Tesla's CEO to an analyst question concerning Tesla's production goals for its Model 3 electric car. Wochos, 985 F.3d at 1191. The analyst asked about “the biggest challenges or bottlenecks in ramping production to 5,000 vehicles per week,” and Tesla's CEO responded by discussing what the company had learned from its Model X car. Id. He explained that the Model X was too complicated to produce due to the company's attempt to include “every cool thing you can imagine all at once.” Id. In his words, that made for a “terrible strategy,” and Tesla had learned from the experience “to start off simple” and to design the Model 3 “to be easy to make.” Id. Even though the comment could be viewed as a statement about the present circumstances of the Model 3's design, Wochos nonetheless held that it was forward-looking. Id. at 1192-93.
The Wochos court did not separately expound on the reasons why the statement about the Model 3's design was forward-looking. Instead, it grouped that statement with others and held that all of them, except for one not relevant to the current analysis, were forward-looking. Id. at 1190-93. In explaining its ruling as to that group, the Wochos court contrasted “subsidiary premises about how various future events will play out” en route to meeting an objective, with “concrete factual assertion[s] about a specific present or past circumstance [that] goes beyond . . . the articulation of predicate assumptions, because it describes specific, concrete circumstances that have already occurred.” Id. at 1192. It explained that the former are forward-looking statements “of the assumptions underlying or relating” to an objective while the latter fell outside the safe harbor. Id. (quoting 15 U.S.C. § 78u-5(i)(1)(D)).
Two observations emerge, providing guidance to this Court on how to assess statements about lessons learned. First, consistent with Lead Plaintiffs' argument, the “lessons learned” statement in Wochos was made in response to a question about future objectives. Id. at 1191. The presence of a question about the future enables a court to distinguish between (A) a forward looking assumption about how future events will play out and (B) a statement of present fact untethered to any future objective. Second, the statement in Wochos identified vague, generic lessons that conveyed little about how Tesla designed its Model 3. As the Ninth Circuit emphasized, an “articulation of predicate assumptions” crosses over into an actionable factual assertion only when it “describes specific, concrete circumstances.” Id. at 1192. The nebulous learnings “to start off simple” and to design the Model 3 “to be easy to make” do not describe concrete circumstances and convey only Tesla's assumption that unspecified design changes would streamline future production. See id. at 1191. More specific lessons, on the other hand, could provide concrete descriptions of what Tesla actually changed and therefore be actionable as a statement about the past or present.
Applying that guidance to the statements in the instant case, the Court concludes that some, but not all of the “lessons learned” statements here are forward-looking. Each of Statements 2, 3, and 9-17 were made in response to questions about future objectives for Intel's 7nm launch or product development cadence. Elliott Decl., Ex. 7 at 11 (Statements 2 and 3); Ex. 10 at 5-6 (Statements 9-12); Ex. 11 at 2-3 (Statements 13-17). Statements 2, 3, 9, and 11-17 are also vague descriptions of Intel's general approach to developing its products, not concrete descriptions of past or present. For instance, Statement 14 identifies a general principle of Intel's approach to designing its 7nm process-to balance scaling and cost with schedule predictability, power, and performance-but it provides no details about any concrete steps that Intel took to achieve that balance. By comparison, Statement 10 offers specific details, declaring that Intel would reduce the complexity of its 7nm designs by foregoing any attempt at 2.4 or 2.7 scaling. Accordingly, Statements 2, 3, 9, and 11-17 are forward-looking while Statement 10 is not.
b. Meaningful Cautionary Language and Actual Knowledge
Statements 2, 3, 9, and 11-17 are assumptions underlying Intel's objectives for its 7nm timeline, so cautionary language that provides meaningful warning about the 7nm timeline also provides meaningful warning about the “lessons learned” statements. In its discussion of statements about the 7nm timeline above, the Court already found that the cautionary language accompanying Statements 2, 3, 9, 11, and 12 is meaningful, so those statements are protected by the PSLRA safe harbor.
Statements 13-17 are not accompanied by cautionary language. See supra n.7. Nonetheless, for the reasons given in the Court's discussion of scienter below, Lead Plaintiffs have not pleaded actual knowledge, so the statements are protected by the safe harbor.
3. Statements Regarding Intel's Status as an IDM
Like the statements about lessons learned, the group of statements about Intel's status as an IDM contain a mix of forward-looking and non-forward-looking statements. Statement 8 about Intel “expect[ing] its IDM model to be intact for the foreseeable future” is plainly a forward looking statement about future plans. See 15 U.S.C. § 78u-5(i)(1)(B). Statements 19 and 20, which both discuss Intel's relationships with external foundries and the value of the IDM model, are also forward-looking because they deal with Intel's expectations “going forward.” Compl. ¶ 157; 15 U.S.C. § 78u-5(i)(1)(B). However, Statements 4 and 5 describe Intel's present relationship with external foundries, Statement 6 is about Intel's current capacity building efforts, and Statement 23 is the unequivocal present assertion, “We are an IDM.” Compl. ¶¶ 136-37, 167. These statements are therefore not forward-looking and fall outside the PSLRA safe harbor.
Because Statements 8, 19, and 20 are forward-looking and Lead Plaintiffs have not pleaded actual knowledge, as the Court discusses below, the statements are protected by the PSLRA safe harbor.
B. Misrepresentation or Omission
Under Section 10(b) and Rule 10b-5, it is “unlawful . . . to make any untrue statement of a material fact or to omit to state a material fact necessary in order to make the statements made, in light of the circumstances under which they were made, not misleading.” In re Cutera, 610 F.3d at 1108 (cleaned up) (quoting 17 C.F.R. § 240.10b-5(b)). Thus, a plaintiff can state a claim by pleading either an affirmative misrepresentation or a materially misleading omission. Wochos, 985 F.3d at 1188.
The PSLRA requires a plaintiff proceeding with a misrepresentation theory to plead the falsity of an alleged misstatement with particularity. Zucco Partners, LLC v. Digimarc Corp., 552 F.3d 981, 990-91 (9th Cir. 2009). This is an “exacting requirement[],” necessitating “‘specific facts indicating why' the statements at issue were false.” Kipling, 2020 WL 2793463, at *14 (quoting Metzler, 540 F.3d at 1070). To plead an omissions theory, a plaintiff must plead facts showing that a statement “affirmatively create[s] an impression of a state of affairs that differs in a material way from the one that actually exists.” Brody v. Transitional Hosps. Corp., 280 F.3d 997, 1006 (9th Cir. 2002). Federal securities laws “do not create an affirmative duty to disclose any and all material information,” though. Matrixx Initiatives, Inc. v. Siracusano, 563 U.S. 27, 44 (2011). Statements are not actionable merely because they are incomplete, and “[o]ften, a statement will not mislead even if it is incomplete or does not include all relevant facts.” Brody, 280 F.3d at 1006.
1. Statements Regarding Intel's 7nm Development Timeline
Each of the challenged statements regarding Intel's 7nm development timeline is protected by the PSLRA safe harbor, so the Court does not assess the falsity of those statements. Lead Plaintiffs argue, though, that forward-looking statements which omitted material past or present facts are not protected by safe harbor. Opp'n at 10, 13. Several district courts in this circuit have agreed. See, e.g., Mulderrig v. Amyris, Inc., 492 F.Supp.3d 999, 1021 n.15 (N.D. Cal. 2020) (collecting cases); Loftus v. Primero Mining Corp., 230 F.Supp.3d 1209, 1225 (C.D. Cal. 2017) (collecting cases). But others differ. See, e.g., Melot v. JAKKS Pac., Inc., No. LA CV13-05388 JAK (SSx), 2016 WL 6902093, at *24 (C.D. Cal. Nov. 18, 2016) (finding that the PSLRA safe harbor protects forward-looking statements alleged to constitute a misleading omission); see also In re Pivotal, 2020 WL 4193384, at *15 (remarking that district courts disagree on whether the safe harbor can protect omissions of historical fact). The Court does not undertake to address that disagreement now, because Lead Plaintiffs have not shown that Defendants made material omissions.
Lead Plaintiffs contend that Defendants' statements about Intel's 7nm development timeline omitted two material facts: (1) that Intel's internal road maps for its 7nm development had changed significantly by December 2019, and (2) Intel missed its hard tape out deadline for its initial 7nm product as of March 2020. Opp'n at 13. The first alleged omission is not one of present or past fact. Intel's internal road maps are forward-looking because they project future product development milestones. For this reason, failing to disclose those road maps does not create “an impression of a state of affairs that differs in a material way from the one that actually exists.” Brody, 280 F.3d at 1006 (emphasis added). At most, it creates an alleged misimpression about how future events will play out. Moreover, Lead Plaintiffs' argument about Intel's road maps is not so much an omissions theory as it is a theory about the falsity of statements regarding Intel's 7nm timeline. To say that Intel's internal road maps had changed is no different than saying public statements about Intel's 7nm development timeline were false, and a plaintiff cannot circumvent the PSLRA safe harbor by simply arguing that a defendant omitted to say that its statements were false.
The omissions theory regarding Intel's hard tape out deadline is equally unavailing. “[C]ompanies do not have an obligation to offer an instantaneous update of every internal development, especially when it involves the oft-tortuous path of product development.” Weston Family P'ship LLLP v. Twitter, Inc., 29 F.4th 611, 620 (9th Cir. 2022). Yet that is precisely what Lead Plaintiffs demand here by arguing that Defendants were obligated to disclose the allegedly missed tapeout deadline.
As such, Lead Plaintiffs have failed to plead actionable omissions related to statements about Intel's 7nm development timeline.
2. Statements Regarding Lessons Learned
All statements regarding lessons learned are forward-looking except for Statement 10, so the Court considers the falsity of only Statement 10. In Statement 10, Swan asserted that Intel was “not going to try to do 2.4 scaling or 2.7 scaling.” Compl. ¶ 153. The complaint contains no allegations about Intel's scaling, so Lead Plaintiffs have not pleaded that Statement 10 is actionably false.
As to the forward-looking statements regarding lessons learned, Lead Plaintiffs again argue that omissions of present or past fact are not covered by the PSLRA safe harbor. Opp'n at 10. They contend that all statements regarding lessons learned, including Statement 10, are misleading omissions for failing to disclose that Intel learned its manufacturing processes were inadequate, that Intel would need to outsource production of its 7nm products, and that Intel was designing its 7nm chips to be outsourced. Id. at 9. Citing to Schueneman v. Arena Pharmaceuticals, Inc., Lead Plaintiffs maintain that a defendant who “tout[s] positive information to the market” must then “disclos[e] adverse information that cuts against the positive” representations. 840 F.3d 698, 706 (9th Cir. 2016) (cleaned up).
Lead Plaintiffs seem to imply that, once a company makes a positive statement, it has an obligation to also disclose every fact that cuts against the positive news. Schueneman does not go quite so far. If Lead Plaintiffs were correct, the federal securities laws would essentially function as an obligation to make complete disclosures of adverse information, but the Ninth Circuit has “expressly declined to require a rule of completeness for securities disclosures.” Intuitive Surgical, 759 F.3d at 1061; see also Brody, 280 F.3d at 1006 (“We conclude that neither Rule 10b-5 nor Section 14(e) contains a freestanding completeness requirement.”). Indeed, a close read of Schueneman reveals that its holding was more limited. It held that a company touting positive information must “do so in a manner that wouldn't mislead investors,” and said only that disclosing adverse information was one way to avoid misleading investors. Schueneman, 840 F.3d at 706. Further, Schueneman emphasized that the statements at issue were misleading because the defendants “affirmatively represented that ‘all the animal studies that [had] been completed' supported . . . approval” of a new drug, but the defendants knew of one study that presented significant problems. Id. at 707-08 (alteration in original). Nothing in Schueneman changes the rule that an omission is actionable only if it “affirmatively create[s] an impression of a state of affairs that differs in a material way from the one that actually exists.” Brody, 280 F.3d at 1006.
In contrast to the statements in Shueneman, none of the “lessons learned” statements here create an affirmative misimpression. Most were vague assurances about how Intel would do better. For example, Swan told investors that Intel was focusing on “lessons learned coming out of the challenges we had with 10 and how we're capturing those lessons learned as we think about the next 2 generations,” Compl. ¶ 134 (Statement 2), or that “good news is we feel like we've got [it] fairly well dialed in” based on learnings from the 10nm process. Id. ¶ 153 (Statement 9). Even more specific statements, like Renduchintala's statement that “one of the most important lessons is to make it easy and fast for our design teams to be able to migrate through intra-node transitions,” Id. ¶ 156 (Statement 17), do not remotely touch on the subject matter of the alleged omissions. In the absence of any reference to manufacturing capabilities or outsourcing, there is no reason to believe that an investor would form any understanding about those topics from the “lessons learned” statements, let alone that Defendants created an affirmative misimpression. Lead Plaintiffs' reliance on City of Sterling Heights General Employees' Retirement System v. Hospira, Inc., No. 11 C 8332, 2013 WL 566805 (N.D. Ill. Feb. 13, 2013), does not change the analysis. Despite one defendant stating that the company had “taken our learnings” and applied them to manufacturing operations, the court never addressed that statement when analyzing the alleged omissions. Id. at *6.
Accordingly, the Court finds that Lead Plaintiffs have failed to plead that the “lessons learned” statements are actionable misstatements or omissions.
3. Statements Regarding Intel's Status as an IDM
Lead Plaintiffs contend that each of the statements regarding Intel's status as an IDM is a misleading omission. They argue Defendants failed to disclose that Intel was planning to outsource 7nm production to external foundries and had been designing its 7nm chips to be outsourced. Opp'n at 10-11. Their arguments are not persuasive.
Lead Plaintiffs have not pleaded facts showing that Intel had definitive plans to outsource production of its 7nm chips when Defendants made the challenged statements about Intel's IDM model. They rely on Intel's Q2 2020 earnings call in July 2020, where Swan revealed that Intel created “contingency plans” to outsource production of Intel's 7nm chips if there were difficulties with its process. Compl. ¶ 104. This, they maintain, demonstrates Intel's intent to “withdraw[] from [its] IDM model for its leading-edge products for the first time in history.” Opp'n at 11. Contrary to Lead Plaintiffs' assertion, the existence of a contingency plan does not evince an intent to execute that plan. By its nature, a contingency plan is a last resort, and a company making contingency plans would prefer not to implement them unless forced to do so. Intel's contingency plan to utilize external foundries for production of its 7nm products represents merely the possibility that Intel might do so in the future. The omission of such a future possibility does not create an impression of Intel's present state of affairs that differs from the one that exists, and Lead Plaintiffs cite no authority implying otherwise. See Brody, 280 F.3d at 1006. They point to SEB Investment Management AB v. Align Technology, Inc., 485 F.Supp.3d 1113, 1131-32 (N.D. Cal. 2020), for the proposition that it is misleading to conceal a significant change in business practices. But SEB did not address a situation where only the possibility of change was omitted; by the time of the challenged statement in SEB, the company there had already implemented changes in its business practices. Id. Sjunde AP-Fonden v. General Electric Company, 417 F.Supp.3d 379 (S.D.N.Y. 2019), is also in apposite. The alleged concealment in Sjunde involved Item 303 of SEC Regulation S-K, Id. at 407-09, a regulatory violation which Lead Plaintiffs do not allege here.
And Lead Plaintiffs have not identified the point at which the possibility of outsourcing morphed into a decisive plan. They suggest that, because designing chips for manufacture in external foundries would take eight to twelve months, Intel must have made the final call to outsource several months before Swan announced the decision in July 2020. Opp'n at 23; Compl. ¶ 106. But that inference rests on several assumptions unsupported by the allegations of the complaint. First, it assumes that Intel would not have begun designing its 7nm chips for external production until after the final decision to utilize external foundries. Yet, nothing in the complaint explains why Intel would not have designed its 7nm chips for both internal and external manufacture from the start. Doing so would be consistent with the allegations that Intel had contingency plans in place because it would allow Intel to shift gears with minimal delay once a final decision was made. Second, even assuming that redesign work commenced only after a final decision to outsource, it does not follow that the decision had been made as of the time of Intel's challenged statements about its IDM model. The latest of Intel's IDM statements came on January 24, 2020 in Intel's 2019 Form 10-K. Compl. ¶ 167 (Statement 23). If Intel made the decision to outsource shortly after that, and redesign took the maximum twelve months, the designs would still be ready by February 2021. That would leave approximately one year before Swan indicated Intel's first 7nm product would be released in late 2021 or early 2022. Id. ¶ 103. Though Lead Plaintiffs suggest this is impossible, their complaint does not explain why that is so.
Finally, the allegation that Intel had already been designing its leading-edge chips for external manufacture a “couple of years” before Swan disclosed Intel's outsourcing plans does not change the analysis. Id. ¶ 106. The Court explained above that such design work is fully consistent with Intel's alleged contingency plans. Just as it was not misleading for Defendants to omit those contingency plans, it was not misleading for Defendants to omit their preparations for those plans.
4. Statement Regarding Keller's Departure
Lead Plaintiffs allege it was false and misleading for Intel to state that Keller was leaving for personal reasons because he actually left due to disagreements over Intel's 7nm process. Opp'n at 16. But they have failed to plead that the statement is a material misstatement or omission. The complaint contains no facts showing that personal reasons were not a contributing factor, and far from the statement creating an affirmative impression that all was well with Intel's 7nm process, Lead Plaintiffs allege that analysts and investors reacted with immediate skepticism and concern about Intel's product development efforts. Compl. ¶ 96. Courts have found similar statements evoking similar responses by analysts to not be actionable under Rule 10b-5, In re Fox hollow Techs., Inc. Sec. Litig., 359 Fed.Appx. 802, 805 (9th Cir. 2009), and the Court likewise finds that Statement 28 about Keller's departure is not actionable.
C. Scienter
A plaintiff bringing securities fraud claims must allege facts establishing a strong inference of scienter. Tellabs, 551 U.S. at 324. Scienter can be established by showing either an “intent to mislead investors” or deliberate recklessness. Glazer Cap. Mgmt., L.P. v. Forescout Techs., Inc., __ F.4th __, 2023 WL 2532061, at *9 (9th Cir. Mar. 16, 2023) (quoting In re NVIDIA Corp. Sec. Litig., 768 F.3d 1046, 1053, 1059 (9th Cir. 2014)). “Deliberate recklessness is a higher standard than mere recklessness and requires more than a motive to commit fraud.” Id. Instead, it is “an extreme departure from the standards of ordinary care[,] which presents a danger of misleading buyers or sellers that is either known to the defendant or is so obvious that the actor must have been aware of it.” Webb v. Solarcity Corp., 884 F.3d 844, 851 (9th Cir. 2018) (alteration in original) (quoting City of Dearborn Heights Act 345 Police & Fire Ret. Sys. v. Align Tech., Inc., 856 F.3d 605, 619 (9th Cir. 2017)). Deliberate recklessness “only satisfies scienter under § 10(b) to the extent that it reflects some degree of intentional or conscious misconduct.” NVIDIA, 768 F.3d at 1053 (quoting In re Silicon Graphics Inc. Sec. Litig., 183 F.3d 970, 977 (9th Cir. 1999)).
When considering scienter, a court considers “all reasonable inferences to be drawn from the allegations, including inferences unfavorable to the plaintiffs.” Gompper v. VISX, Inc., 298 F.3d 893, 897 (9th Cir. 2002); see also Tellabs, 551 U.S. at 323-24. To satisfy the PSLRA, an inference “must be more than merely ‘reasonable' or ‘permissible'”-the inference must be “cogent and at least as compelling as any opposing inference one could draw from the facts alleged” after considering all allegations holistically. Tellabs, 551 U.S. at 322-24. Further, a plaintiff who uses statements from confidential witnesses to demonstrate scienter is required to describe those witnesses with “sufficient particularity to establish their reliability and personal knowledge.” Zucco, 552 F.3d at 995. If a witness's statement is corroborated by other factual information, the plaintiff need not name its sources. Id. But if there is no corroborating information, “the complaint must provide an adequate basis for determining that the witnesses in question have personal knowledge of the events they report.” Id.
1. Statements Regarding Intel's 7nm Development Timeline
a. News Reports
Leads Plaintiffs first argue that news reports from Demerjian and Wccftech show that Defendants were aware of problems with Intel's 7nm chip efforts. Opp'n at 17-20. They identify two reports by Demerjian: (1) a December 12, 2019 report that Intel had pushed back its road maps for certain 7nm products to the second half of 2023, Compl. ¶ 77, and (2) a July 24, 2020 report that Intel had missed its hard tape out deadline on March 31, 2020. Id. ¶ 86. They also identify a Wccftech article discussing leaked slides that allegedly indicated certain 7nm products would not come to market until the second half of 2023. Id. ¶¶ 91-92.
As an initial matter, the parties vigorously contest whether and to what extent the Court is obligated to assess the reliability of these articles. Defendants argue that the Zucco standard for confidential witnesses should also apply to anonymous sources cited in public articles and reports. Mot. at 15. Lead Plaintiffs respond that the Zucco test applies only to confidential witnesses and that a court in this district had previously held the test to be inapplicable to reports based on anonymous sources. Opp'n at 17-18 (citing In re Apple Inc. Sec. Litig., No. 19-cv-02033-YGR, 2020 WL 6482014, at *11 (N.D. Cal. Nov. 4, 2020)).
The Court disagrees with Lead Plaintiffs to the extent they suggest that there is no place for a court to assess the reliability of news reports at all. The reason a plaintiff is required to plead the reliability of confidential witnesses is to prevent her from “set[ting] forth a belief that certain unspecified sources will reveal, after appropriate discovery, facts that will validate her claim.” Apple, 2020 WL 6482014, at *11 (quoting Silicon Graphics, 183 F.3d at 985). That rationale applies with equal force to both confidential witnesses and anonymous sources in news reports, both of which are “unspecified sources.” Indeed, the court in Apple did not wholesale reject the notion that courts should assess the reliability of news reports. It held only that a plaintiff need not always provide “particularized descriptions of [anonymous sources] to establish their personal knowledge of alleged facts.” Id. at *11 (citing Zucco, 552 F.3d at 995). That standard is what Zucco requires if there is no additional factual information corroborating an unknown source's statements, but when there is corroborating information, a plaintiff need not offer such details. Zucco, 552 F.3d at 995. Instead, in the latter situation, a court determines whether those sources are likely to have relevant personal knowledge by reference to multiple factors, including “the level of detail provided by the confidential sources, the corroborative nature of the other facts alleged . . ., the coherence and plausibility of the allegations, the number of sources, [and] the reliability of the sources.” Id. (citation omitted). The Apple court performed precisely this second analysis, determining that it could credit allegations based on reports that were corroborated by formal announcements from Apple's suppliers. 2020 WL 6482014, at *11.
At the same time, the Court does not agree with Defendants that a plaintiff need always describe anonymous sources in news reports with particularity, a position which the Apple court rejected. Id. Though a source's reliability is not corroborated by the sole fact of being “referenced in the newspaper,” In re Wet Seal, Inc. Sec. Litig., 518 F.Supp.2d 1148, 1172 (C.D. Cal. 2007), the fact that a source is cited by a media outlet carries weight. For example, In re McKesson HBOC, Inc. Securities Litigation held that, “if [a] newspaper article includes numerous factual particulars and is based on an independent investigative effort, it is a source that may be credited.” 126 F.Supp.2d at 1272. Thus, the Court finds that it must assess the reliability of news reports by determining either that the reports contain particularized descriptions of anonymous sources or that the media outlet responsible for the report is reliable under the Zucco factors.
Applying that test to the articles cited by Lead Plaintiffs, the Court concludes that Demerjian's July 24, 2020 article, and the Wccftech article cited by Lead Plaintiffs, are not reliable. Demerjian's July 24, 2020 article about Intel's tape out deadline neither described its sources nor was corroborated by any other allegations. See Elliott Decl., Ex. 15. The Court accordingly does not consider that article. And the Wccftech article, although factually corroborated by other sources, is enveloped in too many indicia of unreliability for the Court to credit. The article relies on slides of indeterminate provenance, partially in Russian, that were first posted on Twitter by an unknown and unnamed leaker. Compl. ¶¶ 90-92; Elliott Decl., Ex. 18. What is more, the article freely acknowledges that it found the slides on Twitter, and nothing in the article demonstrates any efforts to independently verify the authenticity of those slides. Elliott Decl., Ex. 18. Under these circumstances, the Court cannot treat the article as a reliable source.
On the other hand, the Demerjian's December 12, 2019 article is reliable. It stated that one of Intel's 7nm CPU products, known as Granite Rapids, had been delayed approximately one or two years. Compl. ¶ 77. Although it does not describe its sources, see generally Elliott Decl., Ex. 14, its conclusions are corroborated by other allegations. In particular, as of at least December 2019, FE 1 was allegedly told that 7nm products were 1-2 years behind schedule, and Intel announced a similar timeline when it disclosed delays to its 7nm CPU products. Compl. ¶¶ 78, 102. In addition, Lead Plaintiffs pleaded facts showing that Demerjian was respected by analysts and journalists covering the semiconductor industry, and that he had been responsible for breaking several major news stories in the industry. Id. ¶¶ 75-76. In response, Defendants flag that the Ninth Circuit previously criticized Demerjian's articles as “secondhand.” NVIDIA, 768 F.3d at 1058. In this instance though, given that Demerjian's information has been corroborated by an internal Intel source, the Court credits his December 12 article.
Still, Demerjian's December 12 article does not support a finding of scienter. For one, the article speaks to delays with Granite Rapids, which is Intel's 7nm CPU product. By contrast, most of the challenged statements about Intel's 7nm timeline are about Intel's first 7nm product, a GPGPU product called Ponte Vecchio. Compl. ¶ 62; see, e.g., Id. ¶ 132 (Statement 1: “We are on track to launch our first 7-nanometer-based product, a data center-focused discrete GPU, in 2021.”), ¶ 153 (Statement 12: “[W]e feel pretty good about . . . launching our first 7-nanometer product in the fourth quarter of 2021.”), ¶ 187 (Statement 30: “[Intel's] 7-nanometer process ‘remains on track' with first products due by the end of 2021.”). Intel's CPU products were always intended to come out after Ponte Vecchio, Id. ¶ 63, so Demerjian's December 12 article cannot support scienter as to statements regarding Intel's first 7nm product. More generally, there is no indication that the delays discussed in the article were ever communicated to any of the Individual Defendants, and in fact, Demerjian explicitly wrote that it was possible “Intel's top management [were] so untethered from what is happening at the engineering level” that they were unaware of the delays. Id. ¶ 77. The fact that Defendants mentioned “roadmaps” in some public statements, e.g., Id. ¶ 172 (Statement 26), without more, does not support an inference that Defendants had access to the road maps that Demerjian described. “Roadmap” is a generic term that could refer to Intel's goals rather than a specific document, and the complaint offers no reason to believe there was a single, unified roadmap that the entire company and all its employees operated off of. Absent any suggestion that information about delays was passed upwards or made available to Individual Defendants, the Court cannot conclude that they had actual knowledge or were deliberately reckless in ignoring the delays.
b. Confidential Witnesses
Lead Plaintiffs contend that allegations from FE 1 and FE 2 also demonstrate scienter. Opp'n at 18, 22. Defendants disagree, arguing that FE 1 is not reliable and FE 2 is irrelevant. Mot. at 17-18.
The allegations from FE 2 may be quickly set aside. FE 2 notes only that there were problems with 7nm yield without providing any detail on when those yield issues were observed or how they would affect the 7nm schedule. Compl. ¶ 79. The bare allegation that there were yield problems at some unspecified time does not support an inference that any Individual Defendant knew of delays or was deliberately reckless.
FE 1's allegations do contain time frames. FE 1 observed that there were 7nm delays at least as of December 2019, and FE 1 further stated that Keller informed Swan and Intel's Board of Directors about problems with 7nm in May or June of 2020. Id. ¶¶ 78, 93-94. Yet, neither allegation supports scienter. The first allegation regarding delays in December 2019 is flawed for some of the same reasons that Demerjian's December 12 article was flawed: There is no indication that the information from FE 1 made its way to any Individual Defendant. See id. ¶ 78. The second allegation about Keller's conversations with Swan and the Intel Board suffers from hearsay issues. “[T]he fact that a confidential witness reports hearsay does not automatically disqualify his statement from consideration in the scienter calculus.” Zucco, 552 F.3d at 997 n.4. But hearsay “may indicate that a confidential witness['s] report is not sufficiently reliable.” Id. Applying this standard, the Zucco court declined to consider allegations from confidential witnesses when they involved multiple layers of hearsay. Id. at 997. That is the same situation here, where FE 1 was told by Intel's former VP of Marketing, who was told by Keller, what Keller allegedly told Swan and the Intel Board. Compl. ¶¶ 93-94. The Court finds that FE 1's allegation based on this chain of hearsay is not sufficiently reliable to credit.
Defendants also argue that FE 1 is generally not reliable under Zucco because FE 1 was a marketing analyst who had no reason to know about Intel's product development efforts. Mot. at 17-18. As Lead Plaintiffs explained in the complaint, though, marketing employees needed to be apprised of product time lines so they could communicate with customers. Compl. ¶ 78. That is sufficient explanation to satisfy Zucco.
In sum, the Court determines that Lead Plaintiffs' confidential witness allegations support neither actual knowledge nor deliberate recklessness.
c. Core Operations
As a third argument for scienter, Lead Plaintiffs invoke the core operations inference. Opp'n at 20-22. They assert that Intel's 7nm process was so important to Intel that it would be absurd for Individual Defendants to not be aware of delays. Id. at 20-21. They also argue that Individual Defendants had access to Intel's 7nm road maps and specifically admitted to monitoring 7nm development. Id. at 21. Defendants answer that generic allegations of “monitoring” 7nm development does not meet the high bar required for a core operations inference. Mot. at 21-22; Reply at 13.
The Court may consider the core operations theory in three circumstances. First, allegations about core operations “may be used in any form along with other allegations that, when read together, raise an inference of scienter.” S. Ferry LP, No. 2 v. Killinger, 542 F.3d 776, 785 (9th Cir. 2008). Second, such allegations may satisfy the PSLRA if “they are particular and suggest that defendants had actual access to the disputed information.” Id. at 786. Finally, “in rare circumstances where the nature of the relevant fact is of such prominence that it would be ‘absurd' to suggest that management was without knowledge,” bare allegations of involvement in core operations “without accompanying particularized allegations” can establish scienter.” Id.
Beginning with the third method of applying the core operations theory, the Court determines that this situation is not the kind of “rare circumstance” where it would be absurd for Individual Defendants not to know the exact progress of Intel's 7nm development. Lead Plaintiffs have alleged that Intel's 7nm development was of great importance to the company and to investors. Compl. ¶¶ 41-53. South Ferry's third option, though, requires more. In most securities fraud cases, the topics about which a company allegedly misled the market will be important to the company and to investors, so importance, without more, is not the “rare circumstance” envisioned by South Ferry. See Gammel, 905 F.Supp.2d at 1078 (“[I]t does not automatically follow from the ‘core' nature of HP's PC and printer businesses . . . that each Individual Defendant was immediately aware of developments in HP's [] strategy.”). None of Lead Plaintiffs' authorities compel the opposite result because none of them found that the core operations theory was sufficient on its own to establish scienter; in each case, the inference of scienter was bolstered by extensive supporting allegations. See Thomas v. Magnachip Semiconductor Corp., 167 F.Supp.3d 1029, 1045 (N.D. Cal. 2016) (considering other allegations, including an admission of fault); Hatamian v. Advanced Micro Devices, Inc., 87 F.Supp.3d 1149, 1162-64 (N.D. Cal. 2015) (relying on confidential witness allegations).
The Court also finds that the second method of applying the core operations theory does not support scienter. Lead Plaintiffs focus on statements by Swan and Renduchintala that “we monitor” 7nm development. Opp'n at 21 (citing Compl. ¶¶ 57, 191). But an acknowledgement that “we” monitor certain developments does not rise to the level of particularized allegations of access. In context, the use of “we” does not suggest that either Swan or Renduchintala personally monitored 7nm development so much as it suggests Intel did so as a whole. The Court does not find that these allegations support scienter. Where such allegations have proven sufficient to establish scienter, they involved the tracking of specific metrics. See Shenwick v. Twitter, Inc., 282 F.Supp.3d 1115, 1147 (N.D. Cal. 2017) (monitoring of “DAU” metric); Bielousov v. GoPro, Inc., No. 16-cv-06654-CW, 2017 WL 3168522, at *6 (N.D. Cal. July 26, 2017) (tracking of inventory in the channel). In those cases, there was no question about exactly what was being tracked and what defendants would know. In comparison, allegations that Swan and Renduchintala monitored 7nm progress provide no particularized details about what information they would have been privy to or why access to that information supports scienter. See Fadia v. FireEye, Inc., No. 14-cv-05204-EJD, 2016 WL 6679806, at *16 (N.D. Cal. Nov. 14, 2016) (“At a minimum, Plaintiffs needed to have provided information about . . . which facts the Defendants were exposed to, and why this exposure supports an inference of scienter.”).
Finally, the Court will address the first methods of applying core operations in its holistic analysis below.
d. Departures of Swan and Renduchintala
Lead Plaintiffs also appear to argue that Keller's departure supports scienter. Opp'n at 22. To the extent they so argue, the Court finds that his departure is not indicative of scienter because FE 1's account of Keller's departure is not reliable.
Resignations and departures “may in some circumstances be indicative of scienter.” Zucco, 552 F.3d at 1002. To support scienter, a plaintiff “must allege sufficient information to differentiate between a suspicious change in personnel and a benign one.” Id. That is, a plaintiff must allege the departure was “uncharacteristic” or “accompanied by suspicious circumstances.” Id. Here, Swan and Renduchintala left shortly after the 7nm delays were announced, Compl. ¶¶ 117, 128, supporting an inference that they were let go due to failures related to Intel's 7nm development. Such departures are hardly uncharacteristic since “[m]ost major stock losses are often accompanied by management departures.” In re Corner Stone Propane Partners, L.P. Sec. Litig., 355 F.Supp.2d 1069, 1093 (N.D. Cal. 2005). Lead Plaintiffs have not otherwise identified any suspicious circumstances attending to those departures, so on these allegations, the Court finds the inference that Swan and Renduchintala were terminated for performance failures to be more compelling than the inference of scienter urged by Lead Plaintiffs.
e. Holistic Review
The Court closes with the holistic review required by Tellabs, 551 U.S. at 322-23. Taken together, Lead Plaintiffs' allegations paint a picture that there were problems with Intel's 7nm development, and perhaps, given the importance of 7nm to Intel, that the Individual Defendants were generally aware that issues existed. The allegations go no further, though. Because the complaint is lacking allegations describing with particularity the information that Individual Defendants received, or the documents that they had access to, the Court cannot infer that any Individual Defendant knew Intel could not meet its 7nm goals or were deliberately reckless in not realizing.
Defendants also raise affirmative arguments against scienter, contending that a lack of stock sales and the implausibility of Lead Plaintiffs' theory weigh against an inference that Defendants intended to mislead or were deliberately reckless. Mot. at 22-23. They argue that there was no reason for Defendants to conceal product issues because those issues would inevitably come to light, and that absent stock sales, there was no other motive for Defendants to mislead. Id. Lead Plaintiffs respond that absence of motive is not fatal, and in any case, Defendants were motivated to conceal problems with 7nm to forestall a customer exodus. Opp'n at 24-25.
Defendants are correct that “a lack of stock sales can detract from a scienter finding.” Webb, 884 F.3d at 856. But that is not inevitably so. When a theory of scienter is not based on allegations that a defendant would benefit from inflated stock prices, a lack of stock sales does not affect the scienter analysis. In re Splunk Inc. Sec. Litig., 592 F.Supp.3d 919, 949 (N.D. Cal. 2022). Lead Plaintiffs do not allege a theory dependent on stock sales, instead focusing on a motive to delay, Opp'n at 24-25, so the Court does not weigh the lack of stock sales against scienter. The Court does observe, however, that it is not clear why there would be a motivation to delay. Delay is sometimes rational. For example, In re Alphabet, Inc. Securities Litigation found that it was rational for Alphabet to avoid disclosure of cyber security information for the purpose of delay because it wished to avoid attention at a time when there was already public scrutiny of similar issues related to the Facebook-Cambridge Analytica scandal. 1 F.4th 687, 706-07 (9th Cir. 2021). Lead Plaintiffs do not offer a comparably reasonable justification for its theory of customer exodus. There is no allegation or explanation for how customers' belief that a product would be released in the future caused them to do business with Intel at the time of the challenged statements. This problem would not necessarily defeat scienter if Lead Plaintiffs had pleaded a sufficiently strong inference of scienter through other allegations. See Nguyen v. Endologix, Inc., 962 F.3d 405, 415-16 (9th Cir. 2020) (noting that plaintiff's theory of scienter was implausible on its face but then proceeding to assess whether other allegations “surmount[ed] her plausibility problem”). But Lead Plaintiffs have not done so, and therefore the implausibility of their theory weighs against scienter. Id. (finding lack of scienter where plaintiff's theory “does not make a whole lot of sense”).
In conclusion, even on a holistic review, Lead Plaintiffs have failed to plead a strong inference of scienter.
2. Statements Regarding Lessons Learned and Intel's Status as an IDM
Lead Plaintiffs combine their scienter arguments for the “lessons learned” statements and IDM statements because their theories of omission for both sets of statements rest on the outsourcing of Intel's 7nm chips. Opp'n at 22-23. They premise their scienter argument on the idea that, due to the length of time required to redesign Intel chips for external manufacture, Intel necessarily made the decision to outsource by the time the challenged statements were made. Id. That decision, they argue, was so important that only the Individual Defendants, in their roles as senior executives, could have authorized it. Id. As the Court already determined above, Lead Plaintiffs' assumptions about timing do not hold up, so their theory of scienter likewise fails.
In so finding, the Court observed that no challenged statements about Intel's IDM structure were made after January 24, 2020. The same is true of statements about lessons learned.
3. Statement Regarding Keller's Departure
Lead Plaintiffs argue that FE 1's account of Keller's departure, and the temporal proximity of the departure from Intel's announcement that 7nm would be delayed, support a finding of scienter. Opp'n at 23-24. The Court already found that FE 1's account is unreliable for purposes of scienter, and temporal proximity cannot establish scienter by itself. Apple, 2020 WL 6482014, at *10. Thus, Lead Plaintiffs have failed to plead a strong inference of scienter as to Keller's departure.
D. Loss Causation
The Court need only address loss causation if a plaintiff has otherwise pleaded actionable misstatements or omissions. Fadia, 2016 WL 6679806, at *17. Because Lead Plaintiffs have failed to plead any actionable misstatements or omissions, the Court declines to perform a loss causation analysis.
* * *
Lead Plaintiffs have failed to plead actionable misstatements or omissions and have failed to plead scienter. Many of the challenged statements are also immunized under the PSLRA safe harbor. Accordingly, the Court Grants Defendants' motion to dismiss the Section 10(b) claims.
III. SECTION 20(A)
Lead Plaintiffs have failed to state a claim for violations of Section 10(b) and Rule 10b-5, so their Section 20(a) claim must also be dismissed. City of Dearborn Heights, 856 F.3d at 623. The Court therefore Grants Defendants' motion to dismiss the Section 20(a) claims.
IV. LEAVE TO AMEND
A court “should grant leave to amend . . . unless it determines that the pleading could not possibly be cured by the allegation of other facts.” Lopez v. Smith, 203 F.3d 1122, 1127 (9th Cir. 2000) (citation omitted). Here, Defendants argue that dismissal should be with prejudice because Wochos forecloses any possibility of successful amendment. Mot. at 25. But the Court cannot determine that issues surrounding the non-forward-looking statements are unable to be cured by amendment. Nor can the Court conclude that, as to the forward-looking statements, Lead Plaintiffs will be unable to plead facts showing that Intel's cautionary language was not meaningful or that Defendants had actual knowledge of falsity. Because the Court cannot determine that amendment would be futile, it Grants leave to amend.
CONCLUSION
The Court Grants Defendants' motion to dismiss with leave to amend to cure the deficiencies identified in this Order. Lead Plaintiffs shall file their amended consolidated complaint by May 3, 2023.
Lead Plaintiffs do not respond to Defendants' puffery arguments. They therefore concede the point, though the Court has already determined that each of the challenged statements must be dismissed on other grounds. See Ardente, Inc. v. Shanley, No. 07-4479 MHP, 2010 WL 546485, at *6 (N.D. Cal. Feb. 10, 2010) (“Plaintiff fails to respond to this argument and therefore concedes it through silence.”).
IT IS SO ORDERED.