Oral argument — Birnbaum v. General Electric Company

The first case on the calendar is number 20-1741 in Wray General Electric, et al. Counsel, please proceed. Good morning, Your Honors. May it please the Court, my name is Julie Reiser on behalf of the Plaintiff Appellant, Oklahoma Teachers Retirement System. This case is brought under the Exchange Act. The complaint asserts that GE had a duty to disclose the recall of the key components within GE's most important product in order to not mislead investors. As the market learned about the blade oxidation defect in the HA turbine, GE's stock price fell more than 12% over a series of partial disclosures. The stock fell even further on news that GE was taking a $480 million warranty reserve over two quarters, which amounted to 38% of third-quarter profit and 27% of fourth-quarter profit. And GE fell from first to last place in market share for gas turbines at the end of the class period. GE's HA gas turbine was a flagship product because of its promise to revitalize the company. GE marketed the turbine as being able to generate more energy than competitor turbines because its blades could run at higher temperatures, allowing them to spin faster, save millions of dollars in fuel, and generate unprecedented efficiency for power plants. Prior to the class period, though, GE learned through the root cause and financial impact analysis that went to the company's most senior officers that it would eventually have to replace all of the blades on the turbines due to a serious oxidation defect. Defendants also knew that revealing the defect would expose GE's power division's significant operational and reputational vulnerabilities. So, GE intentionally or recklessly concealed the defect for nearly a year in the hope that it wouldn't become public. While GE repeatedly and misleadingly described the HA turbine as meeting and in many cases exceeding performance goals at every customer site today, by contrast, the same senior officers approved a plan to secretly tell customers to run the HA turbines less efficiently, thereby generating less power to avoid cracking and catastrophes. In early September 2018, the defect caused a blade customer preemptively shut down three other HA turbines, but GE downplayed the problem. It said the blade had an oxidation issue that required minor adjustment, and it doubled down on the claim that the HA turbine was reliable. In fact, GE was so concerned that negative information about its crown jewel product and power division would become public that management and the board conducted a mole leaks to Steven Tusa, who was a bearish analyst at J.P. Morgan and had published detailed reports about problems in the power division, including the severity of the turbine defect. As investors belatedly learned of the blade defect and its consequences, for an $80 million reserve for warranty coverage, GE's stock price fell repeatedly and significantly. The GE did not challenge the materiality of the concealed information. Therefore, to find that defendants had a duty to speak about the defect, the court need only determine whether the challenge statement gave a false impression about the performance and reliability of GE's most important product. To quote this court in Abramson... Ms. Reiser? I'm Ms. Reiser. I'm going to interrupt for a minute. This is Judge Carney. I understand that part of the defendants' theory in response is that your claim is really one of fraud by hindsight, that it's, you know, easy to say now that these statements were overly optimistic, but that GE failed only to predict that it's the solution that it had in hand to the oxidation issue, that it would fail. What would you point to as the best way to rebut that claim, that they were not just looking backwards, but that it was well known to them at the time that their solution would not prove effective? Your Honor, in 2015, GE became aware, not just of the oxidation issue, but that it had resulted in a premature blade break, and that's in the complaint at paragraph 87-163. We also alleged, and the district court took as true, that GE conducted a root cause analysis and determined it had to replace blades fleet-wide by the summer of 2017. Not only that, but by the summer of 2017, GE had determined that it should contact customers to tell them to check for the blade breaking or cracking at more frequent intervals and to run the blades at cooler temperatures. What indication is there in the complaint that that approach would fail, that this was not going to be a sufficient way of managing and containing the oxidation problem? The approach was to recall all of the blades. What GE was trying to do was to hope that it could do it quietly with customers instead of publicly by letting investors know that it had a recall issue. What the complaint alleges is that by summer 2017, GE had a plan in place to reach out to customers, and that, therefore, throughout the class period, GE was aware that it would have to recall all of those blades, and that the expense to do so, the reserve it would take for that coverage, was $480 million. This wasn't an unexpected event. In fact, Justin Gabelli, who's an analyst, reports in September of 2018, after the Exelon break, GE had known about this for a year and that they had anticipated the problem. It can't be fraud by hindsight if you are expecting throughout the class period that oxidation will occur and that you will need to recall all the blades. Then, lo and behold, the rest of the world catches up and finds out that that is correct. You have to replace the blades. The inference here to be drawn is far more plausible that GE executives were very concerned about the market learning that they had a defect on their hands with their most important product. Instead of coming clean publicly, they made an intentional choice to work directly with customers to try to replace the blades. But the problem was- You mentioned, and I'm going to interrupt for a second. You mentioned the $480 million charge against reserves. That was, in comparison, let's talk about materiality. That was a comparison to $5.7 billion in revenue during the same quarter, which seemed a possibly immaterial amount. It was obviously substantial, but maybe not material for purposes of the kind of fraud that you allege. What's your response to that? Your Honor, I believe my time has closed up, but if I could answer your question. The complaint alleges that the $480 million was material when taken against the proper metric, which is how GE reported it. The proper metric that GE reported was against net profit. It's two installments of $240 million in the third and fourth quarters that amounted to in the third quarter, and 27% of the net profits in the fourth quarter. So we disagree that comparing it as to revenue is an appropriate comparison when it is a cost, and therefore should come off the top of profit. And that is the basis to say that this was material, as well as the repeated stock declines and the fact that GE's market share fell from first to last. Very good. Thank you. I have a question. Oh, go ahead. Dr. Baum. Following up on Judge Carney's question, if I understand correctly, it was within two months after the Exelon failure of the new... The Exelon failure involved one machine of the new series, and it was within two months of that, within about 50 days of that, that GE disclosed in a 10Q report an intention to immediately take a $240 million reserve charge against losses, disclosing further an intention to have further similar charges in the future, which covers a good $480 or half a billion in reserves. Now, it seems to me that the Exelon failure of a new generation machine at 10,000 hours was something that gave new urgency and new scope to it, and inevitably, they needed to explore what it was that explained the 10,000-hour failure of that Exelon machine, and to reach a conclusion, to announce publicly within less than two months after that failure, an intention to make reserves of $480 million. I find it hard to square that with evidence showing a strong, strong circumstantial evidence of conscious misbehavior or recklessness. Your Honor, I would have the Court look back to the Novak case, and in particular, at page 313, what the Court wrote was, and this is about Ann Taylor Inventory, the complaint provides specific facts concerning the company's significant write-off of inventory directly following a class period, which tends to support the plaintiff's contention that inventory was seriously overvalued at the time the purportedly misleading statements were made. Here, as I was just saying, the defect was known throughout the class period. The solution also was known. It needed a fleet-wide recall, and so, they were already in progress of replacing blades at the time that Exelon occurred. There was the customer meeting in September 12th and 13th where customers were talking about delayed plant openings because their new blade was being diverted to somebody else's current operating HA turbine. So, the fact of knowledge starting 2015 that a blade break not only could occur, but did occur, that the same heat process was used on the turbines for the 9F and the HA, that the defect was fleet-wide, and that they would ultimately have to recall every one of those blades, means that the inference of that $480 million is much stronger to be said that at the beginning of the class period, they knew that that was the likely warranty reserve and that they would have to take that amount if they had to come clean and disclose it publicly. But instead, they made the intentional choice to try to delay in the hope that good news would overtake the bad. And it falls squarely within Novak, SAIC. What are the specific facts alleged in the complaint that show that in 2017, GE recognized that it would need to replace the HA, all the blades of the HA, as well as a problem with the previous generation turbines? So, at JA 49 to 50, we allege that the oxidation defect had been assessed from an engineering and financial standpoint before the summer of 2017. Those are paragraphs 86 to 91. And because of that, GE was going to have to replace and recall all of the HA turbine blades. At JA 201, paragraph 147, we allege that GE secretly was warning customers to run the turbines at cooler temperatures to check for the defect. And so, then, I'm getting you one more citation, one second. We know of the customer meeting that happened September 12th and 13th, where customers were and that's paragraph 110 and 112 of the complaint. So, there's not really any dispute based on the allegations in the complaint that we have pled that GE's actual knowledge began with a blade break in 2015, that that knowledge extended through a root cause analysis and financial impact analysis that went to GE's senior executives by summer of 2017. And that they secretly agreed or they agreed to secretly engage with customers to try to get the replacement plan underway. And the fact that GE told analysts in September of 2018 that they had known of the problem with the defect with the blades and were actively working with customers only underscores that they had been aware of the situation and the need to replace all of these blades. This was not a question of if the oxidation would occur. It was when the oxidation defect would occur. And GE was actively doing what it could to get new blades in before there was a catastrophic break. So, the timing argument that GE is making is a little bit of a red herring because it didn't matter. I mean, it mattered in the sense that the public became aware of a blade shattering and that caused heightened attention to the problem. But the fact of oxidation, the fact of a shortened lifespan for that blade, that was announced immediately when Exxon became public. Ms. Reiser, you'll have three minutes to rebuttal and unless Judge LaValle would like to ask you a question to follow up right now, we'll hear you a little later. Judge LaValle, does that work for you? Nothing more now, thank you. Very good. Mr. Martinez. Good morning, Your Honors, and may it please the Court. Roman Martinez for the GE defendants. Judge Cote's dismissal of this case should be affirmed for two overarching reasons. First, GE's statements here were not false or misleading. The vast majority of them were innocuous statements of corporate optimism that plaintiffs concede were literally true. None of the pre-Exxon statements implied anything about the lifespan of the H-class blades or about how quickly oxidation would affect those blades. And the post-Exxon statements accurately reported the blade break, its fleet-wide impact, and the likely costs. Second, plaintiffs have not satisfied the PSLRA's demanding standards for Sienter. Before Exxon, as Judge Cote found, GE believed that it had a viable solution for the oxidation issue, which was to repair or to replace the blades. Plaintiffs have not made and cannot make any allegation that GE knew this solution would fail. And in fact, GE had good reason to think that this solution would succeed. The 2015 break had come at 22,000 hours of operation, so GE thought it had plenty of time to implement the coatings or the replacements of the blades without undue disruption or cost. They, of course, learned differently after the Exxon incident, but at that point, they then disclosed that they were going to have to move more quickly. Plaintiffs also failed to cite any of the classic indicia of Sienter. There's no confidential informant claiming that GE lied. They don't have any smoking gun documents, and they have no insider stock sales to point nor do they have any sort of overarching coherent theory of why GE would lie. If GE knew, as plaintiffs suggest, that their solution wouldn't work and that the blades would be breaking right and left, that problem would have been impossible to conceal. It would have come to light, and it would have been pointless for GE to try to hide the truth for, you know, by itself just a few more months. Judge Cote was therefore entirely correct. The most plausible explanation for what happened here is that GE believed its solution would succeed, and there was no Sienter. As for the post-Exxon statement— Mr. Martinez, this is Judge Carney. I was concerned about the statements made in 2018 in September after the Exxon power plant suffered a total blade failure, where the statements were made that the minor adjustments that we need to make don't make the HA any less a record-setting turbine—that's fine, General—and they are meeting their performance goals at every customer site today, and again on September 28th, meeting and in many cases exceeding performance goals at every customer site today. This was on the heels of a pretty dramatic failure. How could that be seen as accurate? Sure. Let me take them in reverse order, Your Honor, if that's okay. Starting with the September 28th statement, I think it's important to look at that statement in context, and the full statement appears at page 870 of the Joint Appendix, and I think the key thing to note about that statement is that it comes after a very fulsome five-point disclosure of the oxidation problem, about what the problem is, about the fact that it had a fleet-wide impact, about the fact that GE had a solution that was going to be implemented that it was going to include shutting down temporarily all four of the Exxon units, and it's only at the end, after sort of explaining the problem and what GE knew and what it was doing about that problem, did GE then reiterate the entirely correct statements about the current performance of the turbine. I think in context, I don't think that there was anything false or misleading about GE saying that although we've diagnosed this issue and we have a solution for fixing blades and replacing blades that we're going to implement, that currently the turbines that are operating are operating at record-setting levels, and I don't think that that last part is disputed. I think with respect to the September 28th statement, that's what I would say, essentially that the context and the other statements and disclosures GE had made about the Exxon break and its aftermath, I think, made clear that there was nothing misleading. With respect to the September... How does this... Wait, let me just interrupt, commenting on what you just said. How does disclosure that it would be necessary to shut down the four Exxon units, how does that adequately disclose the need to replace the blades fleet-wide? I don't think it does in and of itself, Your Honor, but I think that GE was very clear throughout this problem that there was a fleet-wide fix that was necessary and that GE confirmed the fleet-wide impact. And I think, you don't have to take my word for it, if you look at the JA61, the complaint itself says that on September 21st, a GE itself, a GE spokesman confirmed that it was going to have a fleet-wide impact. The Second Amendment also says that GE confirmed this in the conference on September 12th and 11th, that's at page 153 of the JA. There's also the Reuters report from September 20th, JA814, that says that GE confirmed the need for a fleet-wide fix. And so, I think GE was very clear throughout this period, not just because it referred to the four Exline turbines, but because it made broader statements that there was going to be a fleet-wide impact. I really don't think there's any credible case that GE was misleading anyone about that. Is it okay- So, you're saying that it was, excuse me, this is Judge Carnegan, just to follow up on Judge LaValle's question. So, I'm understanding you to say that the statement is that they were meeting their performance goals at every customer site today was actually inaccurate, but the problem was disclosed. So, it was kind of except for what we've just told you about. Otherwise, they're doing what we said they would do at every customer site. Is that correct? No, Judge Carnegan, let me clarify that. I think the statements about the nature of the problem here was not that there was some immediate defect in all the turbines that was preventing them from operating. That wasn't the problem. The problem was essentially that what Exline showed GE was that the lifespan of the turbine was shorter than they had anticipated. That didn't mean that the turbines weren't running perfectly well at the time. It just meant that later, they were going to essentially have oxidation issues sooner than GE had anticipated. The nature of these turbine parts and these turbine blades is not that they're put into the engine and they're supposed to last forever. Everyone understands from day one that they're not going to last forever. The question is, how long are they going to last? The 2015 break showed that there was a problem at 22,000 hours. The Exline break in 2018, the game-changer feature of that break, the key point of information that GE learned was that the problem was manifesting itself a lot sooner at less than 10,000 hours. That didn't change the fact that the turbines that were currently operating had not yet operated at their record-setting capacities. I think that's how you would square those statements. That's why the September 28th statement about the fact that our turbines are working well around the world is totally consistent with the idea that we now understand that they have a shorter lifespan. Therefore, we're going to implement a swap-out, a replacement of these turbines faster than we otherwise would have. Judge Carney, I do want to go back to your original question about the September 20th statement as well. About the minor adjustments. I just want to be very clear about what that was referring to. The September 20th statement in the Reuters article said that, and I'm quoting here, the minor adjustments that we need to make did not make the HA any less of a record-setting turbine. They're meeting and exceeding their performance goals at every customer site today. I think that statement is most clearly a reference to the technical adjustment to the turbines that was going to be necessary. If you look at the second amended complaint, it specifically talks about, this is at JA61 and JA134, drawing on materials that appear at JA844 and JA913. The second amendment complaint explains that the technical adjustment to the blades was that GE was going to, quote, revert to an earlier heat treatment process for making the blades. In other words, GE was saying, we see this problem with the Exelon blades. We understand that we need to fortify them against this oxidation problem. The solution is going to be to revert to an earlier heat treatment process, i.e., a heat treatment process that had been used on earlier sets of blades. That's going to fix the problem. In that sense, the technical adjustment to the manufacturing process was minor. They didn't have to redesign the blade. They didn't have to come up with a brand new system for manufacturing the blades. They just were going to revert to an earlier heat treatment process. That's the fact that was alleged in the complaint by my testimony. Thank you. Mr. Martinez, we gave Ms. Reiser some extra time to answer a question, so I'd like to give you a couple extra minutes as well. I'd ask you to address what your adversary said about Cyanter, about the secret customer meetings, the non-disclosures as well, and some of the elements that she pointed to that suggested, in her view, guilty knowledge. Sure. I think on Cyanter, there are four key points that I'd like to make, just very briefly. Number one, I think that there is no allegation that GE knew that the fix that it had in place before Exelon wasn't going to work. So Ms. Reiser today repeatedly said that GE knew beforehand that it had a fix, and the fix was going to be a fleet-wide replacement of all the blades. That's not consistent with what was alleged in the complaint. If you look at the complaint, I think it's at JA-49, the fix that the plaintiffs themselves cited was that either the blades were going to be coated, they were going to be essentially repaired by putting a new coating on them, or they were going to be replaced. I'm not sure that that matters, because either way, the solution that GE had in place was one that was not going to cause a – if that solution had worked, it wouldn't have caused a material problem, because of the fact that the solution would be implemented before the blades started to break. And so customers wouldn't experience the problem, and the solution would be less costly because they could be implemented on a more leisurely replacement schedule. That all went out the window with Exelon, because Exelon showed GE that the replacements had to happen sooner. But of course, GE can't be held responsible for knowing that the Exelon break was going to happen at less than 10,000 hours, because between 2015 and 2018, it thought that it had more time. It thought that it had at least until 22,000 hours, which was the first time they saw the break in the F-Class turbine in 2015. So I think that's the first point. I think the second point is just – I would refer the court to Judge Coates' analysis of this issue at pages 34 to 35. I think she looked very closely at this and concluded that the non-culpable inference offered by GE, essentially what I've been talking about, about GE's belief that it had a plan in place, that that was decidedly more plausible than the inference of Sienter. I think the third point that I would just emphasize is that this is an unusual case in which they don't really – the other side doesn't really have any of the classic signs of Sienter. There's no confidential informant claiming GE knew it was lying, no smoking gun document, no insider stock sales. And as I said earlier, there really isn't a coherent theory of why GE would try to mislead people. It would be pointless to mislead investors and basically pretend like everything was fine if in reality, as plaintiff's theory sort of suggests, GE knew all along that these blades were on the verge of breaking and were going to start breaking because that was going to be a fact that would have been impossible to hide from the market. So why would GE just sort of cut that out? They haven't put forward any coherent theory for why GE would do that because all GE's deception would do is delay the moment of reckoning by at most a couple of months. So there's really – the theory doesn't hang together. Very good. I think we have your arguments. We'll hear the rebuttal. One more question if I may. Oh, sorry. Judge LaValle, go ahead. What is the relationship between $480 million and GE as a whole annual profit? Thank you. Thank you, Your Honor. I think that's an excellent point. The $480 million is 0.4% of GE's overall revenue, which we think is the relevant marker here. The overall revenue of the company was $122 billion in 2018. It's also 0.4% of GE's overall market cap, which was $115 billion in 2018. And I think if you look at it in terms of the profits, I don't think that that number is alleged in the complaint. But if you look at the financial documents, the effect of the $480 million charge was about $0.05 or so on their earnings per share. And the Motley Fool Investors Report that the plaintiffs themselves cite, and it appears at page 850 of the joint appendix, describes – it says that the $480 million would be a, quote, minor issue if it takes just a few cents off of GE's earnings per share. And that's ultimately where that charge came out in light of GE's broader financial performance. Obviously, GE's a huge company. What is $0.05 as a percentage of GE's 2018 per share income? You know, Your Honor, I don't have that number immediately in front of me. So I think it's around – I don't want to speculate because I don't want to get it wrong. But I will – I guess I would just say that the way that the analysts that were looking at it, when they looked at that $0.05 or a few cents, as the Motley Fool Report said, it was treating that as a minor issue. And that – those are the words of the analyst, a minor issue. But didn't – but the stock price fell significantly at the series of disclosures, correct? It's true that the stock price fell. Doesn't that reflect the market? That's right, Your Honor. And what the Motley Fool analyst report at 8.50 of the joint appendix says is that it was saying that, look, if the impact here turns out to be just a few cents of earnings per share, then it's a minor issue. And Motley Fool said that means that the stock overreacted. In other words, that the real effect just – it didn't justify that kind of stock drop. And I think ultimately that's what happened. They fully disclosed the $480 million. It was just a few cents of the earnings per share for that year. And I think, therefore, it's fairly characterized as – Motley Fool characterized it as a minor issue in the context of the broader GE as a whole. Very good. We've kept you past your time, and thank you very much for your argument. We'll hear rebuttal. Thank you, Your Honor. Thank you, Your – excuse me. Thank you, Your Honor. I wanted to touch on three issues, if I could. Number one is this idea that coding for replacement could have been a minor and corrective step. Second, what GE disclosed on September 19th and how it did not give the truth to the market. And third, why Regulation SK was triggered, as well as other scienter allegations. So, first, I would ask the court to look at paragraph 165 of JA-221, and it said that the protective coatings that GE had hoped to use showed cracking at 7,000 hours. In other words, that coating was never going to take the place of replacing blades feet wide. That was a known defect and a known need to replace them. Therefore, the idea that they could have done this in a minor way – and to be clear, under materiality thresholds with the SEC, it's about 5% of net profits would be material, in addition to all these other factors of the stock drop. It means that at the beginning of the class period, GE had to have thought its solution would be under about $20 million, and that somehow over the course of the year, it ballooned to $480 million. And respectfully, that is not as credible when they've known that there is a fleet-wide defect, and they have been in the process of pulling out blades and replacing them with new ones. Secondly, on September 19th, GE came out with a post. It said, this is a minor issue. It did not mention a blade had shattered inside a turbine. It did not mention that the blades had to be replaced. It just said it was a premature oxidation issue. It did not mention that there were a lack of replacement blades, and they were robbing computer to pay Paul, and this was delaying new plant startup. It also did not mention that three turbines had preemptively shut down. And in fact, by September 26th, that number was five preemptively shut down – well, I guess one was injured, and the other four preemptively shut down for blade replacement. So at the time, on September 28th, when defendants said every site is meeting or exceeding performance guarantees, five of those sites were not even in operation. That's 10% of their turbines not even working. Separately, two weeks later, October 12th, GE revealed that 14 of the blades showed signs of cracking. That's paragraph 154, JA 218. So Exelon wasn't anything new. They knew there was a defect. They knew they needed a fleet-wide recall, and they knew that there was premature cracking. The only thing that wasn't known was the public knowing that this was an issue. Reg SK is designed for this sort of process, right, where if there is a known defect, then the company has to go through the process of saying, is this material – I'm sorry, if I could just finish my sentence – is this going to have a material impact? And in fact, it's phrased the opposite way. It says if there's a known trend or uncertainty, that must be disclosed unless management determines the material effect on the registrant's financial conditions, the result is not reasonably likely to occur. So it has a presumption in favor of disclosure. And just like Panther Partners here, the district court erred in using revenue as a way to get We have clearly pled the fact that they had a defect issue, that they knew of that known trend or uncertainty. And I guess, you know, because of the 2015 break, if we wanted to use the term uncertainty, it still would trigger a warning under Regulation SK so that investors knew that this was a possibility. And I guess last, if the court will indulge me, on Scienter, the complaint plausibly alleges that defendants had a motivation to conceal this information and that it would matter to the market. So there are a number of analyst reports that talk about this issue as minimizing franchise value, hurting GE's turbine brand image, noting that the stock price is falling because of the oxidation issue. And so while I appreciate Judge LaValle's questions about GE's overall profitability, this was the unit of GE that was charged with turning around the entire company. And the fact of a defect within these units did cause the stock price to drop, or at least it's alleged, and analysts agreed, that this was market-moving information. Unless the court has further questions for me. Thank you. Thank you very much, Ms. Reiser. Thank you both. We will reserve decision.