Oral argument — Caleb Ho v. Flotek Industries, Inc.

Good morning, Your Honors. May it please the Court, Steve Hubachek, Robin Scaler, on behalf of the appellant. Your Honors, in the end of 2014, and continuing until the Bronte Report revealed its shortcomings in November 9, 2015, the defendants relentlessly promoted their invention, the FracMax software application that was designed to be the solution to the marketing problems that Flotek was encountering with respect to marketing its CNF products, which they described as the hallmark products. They were trying to overcome a reluctant industry and the product was quite expensive, and so they needed to be able to prove its efficacy. But that promotional campaign was based upon multiple falsehoods, all of which were severely recklessly made. First, the defendants said that the data that FracMax relied upon was unadjusted. But in fact, FracMax contained an algorithm whose very purpose was to modify that data to enable well-by-well comparisons. Second, defendants represented that the evidence that FracMax provided was conclusive, in other words, irrefutable. Any reasonable investor hearing that would assume that they had verified the accuracy of that data, but in fact they recklessly failed to do so. Third, they were actually in the process for several months during the class period of replacing the allocation algorithm, which was the one that was responsible for doing the well-by-well comparisons for at least three months, perhaps longer, and yet they never said anything about that to investors. What was that? I'm sorry. I missed that. Yes, Your Honor. What they were doing is that during the class period for several months, at least three, perhaps more, I think you can infer more under this Court's decision in Plotkin, they were replacing the allocation algorithm. That was the portion of the FracMax software that was responsible for doing their well-by-well comparisons. This was a third party that used the algorithm, or was it the defendants? It was the defendants, Your Honor. There was a third party that was providing them. The Texas Railroad Commission, I guess, gets the data reported to them, and they would provide it to Drilling Info, and then Drilling Info would provide it to the defendants to use in their FracMax. What is the standard that you are arguing would apply to reliance on third party data or third party process? Would it be that you're grossly negligent in relying on it, that you're negligent? It seems like it would require more than just negligently relying on it. Or is it deliberately indifferent to? Your argument is that these defendants, the principles of the defendant, had to know, had to know that this data was being publicized and that it was incorrect. As I understand it, that's your argument, and they're, in part, they're claiming, oh, we relied on a third party or software program or some Texas Railroad Commission data. What is the standard that should be applied for a defendant to rely on a third party? When does that become unreasonable and illegal? Well, the standard here is severe recklessness. But if I could address the fundamentals of your question, Your Honor, that really is the position that the defendants have taken in this brief. But what we're focusing on is the defendant's own conduct here, not Drilling Info's conduct, because what happens here is that Drilling Info provides the information to the defendants. The defendants then have a software application that, through an algorithm, changes the leasehold by leasehold data into well by well data. So that's all on the defendants. They're the ones that do that. Drilling Info doesn't do that. So that's number one. Number two, they came out and said that this evidence is conclusive, i.e., irrefutable. So it was their responsibility to verify that their own evidence that they were telling people was irrefutable, and not just once or twice. We're talking over a period of time from November of 2014 to November of 2015, where they said it at least 12 times, I think. And that applies equally to the three individuals involved, Mr. Chisholm, Mr. Walton, and Mr. Schmitz? Yes, Your Honor, although in candor, the case is strongest as to Mr. Chisholm because he's identified as one of the inventors of FRACMAX. So I think the inference of his knowledge of its workings is the strongest. However, the district court found that the fact that this was Flotech's invention was sufficient to impute knowledge of the workings of Flotech to all of the defendants, but it's definitely strongest as to Mr. Chisholm. Is it your position that this data, even though it might have been flawed, doesn't support the claim that CNF was effective? Your Honor, I don't think that maybe some of the data supported that, but the point was does FRACMAX provide new customers with a basis for choosing to buy the CNF product? That was the purpose of it. Because their problem was that their product was difficult to demonstrate its efficacy, number one, and number two, it was expensive. So what was important for them to be able to grow their business was to be able to attract new customers. And the way they wanted to attract new customers was through using this FRACMAX application. So this case is really about the FRACMAX representations that they were making. And in this instance, after the information came out on November 9 of 2015, everyone knew that the FRACMAX data was unavailable. The stock dropped significantly on November 9 when they revealed that. And then on November 10, after Mr. Chisholm provided his explanation, it dropped even more. It was 19% on the 9th and then 38% on the 10th. Did your clients sell at the low price? Your Honor, I don't know off the top of my head when my clients. You don't know? How could you not know that? This is a damage claim. Did they have damages or not? Well, Your Honor, we're here. The only basis for the district court's dismissal was Cyanar. So I'm here prepared to talk about Cyanar, Your Honor. You don't know the value of your case? Your Honor, I personally don't. I didn't handle this case in the district court. I'm responsible. You don't know if they sold low? Your Honor, whether they sold low, I'm sure that they sold low or they held and they lost as a result of that. Now you're sure. Well, it went back up, didn't it? Yes, Your Honor. All right. So if they didn't sell, they're fine. Your Honor, I'm sorry. I honestly don't know the date that they sold. I find that hard to believe. I apologize for that. Now, again, with respect to the Cyanar inference here, I think that this court's decision in Nathanson provides the length in which this works. Does this case hinge on the comparison of two or three wells against four wells? In other words, how many wells are involved in the alleged false claim that this company's, what do you call it, their product was clearly allowing some wells to produce more than others? How many wells are involved in that comparison? I got the impression there's only about seven wells and that what happened was the defendants represented that the wells that weren't using their product weren't doing as well. But they mistakenly made an adjustment because they thought those wells were producing from units with many wells. And so they had to reduce it down to one well per unit to compare it with the wells that were using their product. I mean, is it that simple or is it more complicated? Does it involve . . . I also see a reference. The district judge talked about 80,000 wells. We're not looking at a false representation as to 80,000 wells, are we? Alleged false representation, I should say. Right. I think most of what Your Honor said is correct. It was a small number of wells, and what they did was that they reduced by a large percentage, I think 40%, the actual production of three of the wells that weren't using their product and didn't reduce the one that did use their product. But the point, Your Honor, is that, number one, once they do that, they had to reveal, which they did on November 10 when the stock dropped even more, was that they actually were using an algorithm to adjust the data, which they had been explaining over and over again that they hadn't done. They said that 10 or 12 times as well. And they also had to take back their statements that their evidence was effectively irrefutable. I don't remember anybody accusing them of saying they weren't using an algorithm. They said they weren't using it. It was simpler words than that. They did say on September 11 that they used an algorithmic econometric database. But what they were saying at the same time, and including after September 11, on September 28 and October 5, that their data was unadjusted. But it was adjusted, and they had an algorithm that was specifically to do that. Software has algorithms all the time, but it's the function. But Judge Bennett said, well, everything is adjusted to some extent with an algorithm. Is that right? Maybe you should just tell us, where did Judge Bennett go wrong in not seeing the scienter here? Well, I think where Judge Bennett went wrong is that he didn't focus on the specific allegations that were making before this court today, which was that throughout that entire year, they told people that the data was unadjusted. That proved to be false because, as was revealed on November 9, in fact, it was adjusted very significantly and apparently not always accurately. Number two, they said that this evidence was conclusive. When you say that evidence is conclusive, you're basically saying that it's irrefutable. That creates a duty on your part to make sure that you actually have a basis for your statement. But on November 10, Mr. Chisholm admitted that, in fact, they never verified the results of that data. So on both of those bases, which Judge Bennett didn't focus on, I think that the defendants clearly demonstrated their severe recklessness by, number one, falsely saying that their data was unadjusted, and, number two, severely recklessly saying that their data was irrefutable when they had no basis for saying that. That's basically the basis for our claim in a nutshell right there. In addition, there's also the fact that during the class period, for several months, they actually realized they had to change the allocation algorithm that was performing this particular function. But they never said anything about that to their shareholders, but instead they continued on and made the same representations I've been talking about this morning. They're replacing the allocation algorithm, but they still said that it was conclusive. So how can it be both that they need to replace it and that it's providing conclusive evidence? Basically, if it's not broke, you don't fix it. If it's already providing conclusive evidence, they didn't need to change it. In fact, on November 10th, when they made their announcements where they admitted these various things, they focused on what they hoped to be the accuracy of their new algorithm after they'd spent a year telling everybody that basically their old algorithm was the best thing since sliced bread. So that basically is the focus of the claim here, Your Honors. I believe at one point the defendants in the brief say basically that CNF works, but it doesn't work as great as we thought it did originally. Would you agree with that as a factual proposition? Or do you still contest that it's junk and it was entirely misleading because it's completely ineffective? Your Honor, our case is really about the FRACMAX application as the marketing tool to get the CNF product greater acceptance and particularly greater acceptance as to new customers. That's why, and I think this is largely outside of the record, when the defendants say there wasn't a mass exodus, that's really not in the complaint. But the fact that there isn't a mass exodus really isn't relevant to what the purpose was here. But I think it is relevant to CNTR if it works, but not to the extent that was represented based upon whatever exemplar data they had. That's one thing. If it doesn't work at all, if it is a useless exercise and it's being trumpeted as this great product, doesn't that have to do with the CNTR of the defendants? Well, I think that it could well. But our point here is that they were using this FRACMAX to market it because even if it's a great product but people don't know about it or they're daunted by the price of it or it's very difficult to figure out whether or not there really is an effect on production, then they're not going to buy it. But in order for Flowtech to grow its business, develop new customers, and sell more and more and increase their top line, that was the point. In fact, if you look at all the analysts' comments right after the news came out, they all emphasized the problem with first-time buyers. And here's just one at paragraph 155. Uncertainty about the accuracy of FRACMAX data and the effectiveness of CNF will serve to deter some first-time buyers. That's at paragraph 155. Very similar thing at paragraph 156. Very similar at 158 and 159. This case is all about their ability to market the CNF, and when the information came out in November 9 of 2015 and November 10 of 2015 that everything that they had been saying for the previous year was not true, they lost a significant amount of value from the stock, and the analysts reacted as I just described. Your Honors, unless you have any further questions, I'll reserve the rest of my time. Thank you, sir. Do you have some time on your phone? Mr. Pecht. Your Honors, Jerry Pecht, Norton Rose Fulbright here on behalf of the Apple Lee. I want to get to some of the very good questions that the Court asked and provide a different context to those, but I first want to start with the fact that this is governed by, I think everybody recognizes, the Private Securities Litigation Reform Act, and it sets a very high astringent pleading requirement that the plaintiffs must meet in order to get past the pleading stage of these cases, and they must plead particularized facts demonstrating a strong, cogent, and compelling inference of fraud, and that's what they failed to do. Now, to get to Judge Dennis's question about the particularized facts regarding an error, the particularized facts in this case are that there was an error in three wells, and those three wells are all within a four-mile-square radius of each other, all operated by the same operator, Sabine Oil and Gas, all in south-central Texas, and there was an error in those three wells. They're all on single-well leases. The CEO said it may have been as simple as putting a 2 instead of a 1 in the software code, and you have a glitch in the software. A glitch in the software does not make out a securities fraud class action case. Well, what bothers me a little bit about that argument is if you're going to make a pitch to people to buy your product and use as an example three wells as you compare against, what, six others or five others? I forget what it is. I mean, you should make sure what you're saying is correct. Recheck it maybe. I understand no rechecking was done according to allegations. And, look, I think in hindsight they wish they had rechecked it, but what they had in this case is they had a third-party provider, a doctor, a Ph.D., Dr. Glenn Collins, Texas A&M, statistics and economics, designed a program for the Department of Agriculture, all of which is described in the materials before the court, designed a program for the Department of Agriculture where chemicals were put in the ground and he determined what the production was going to be from the fertilizers in the ground for agriculture, just like chemicals being put into a well and what's the production from those wells. So this is a man who had expertise in providing this kind of data, and no one thought that he would be providing data that was not accurate. It was just sort of incomprehensible that that would happen. But, look, there are, in his defense, there are thousands of lines of code. Well, did he make the mistake or did the defendants make the mistake? He made the mistake. I mean, who changed the comparison from a multiple well district to a single well? It's in the software, and he did the software. There's no allegation here that anyone with the company has the expertise or skill to write software. And, in fact, the allegations are all about Dr. Glenn Collins, the third-party provider who was – and I'll just – and the pages in the record start at 1651 and 1652 in which they describe that. And these three wells are three of 98,000 wells in seven states, the three that had the error in them. Three in 98,000, and that's at record 1493 to 1495. By the way, there was a mention of Drilling Info. Drilling Info is not the third-party provider. Drilling Info is an aggregator of information from the Texas Railroad Commission that becomes available. That's a different – he got that incorrect. So he raises this issue about that the company did say that the information was unadjusted. The information coming in – and you've got 17 different states. Each state has slightly different information from their respective railroad commissions or their commissions. It all comes in on an unadjusted basis. It's not being adjusted for temperature. It's not being adjusted for pressure. It's not being adjusted for anything. But in the state of Texas, that information is on a lease basis. And remember, this is a customer-facing app. You show it to the customers. You show them their data. They know their data. They know what's being produced from their wells. You cannot mislead them about what's being produced from those wells and wouldn't want to any more than you would try to mislead somebody about what they're paying in property taxes because they know what they're paying in property taxes, and they can look it up publicly. So in single-well leases, you can look it up publicly. So the idea that you would engage in a fraud to mislead your own customers about their own data is bizarre. But that's what's happening. You're showing them their own data, and that data, as they know in the state of Texas, is reported by the Texas Railroad Commission on a lease basis, and it is unadjusted as reported in FRACMAX on a lease basis. But now it's a computer program. It's an app. What do these do? They analyze the data. They analyze the data. And if you want to see the data on an individual well basis, then there has to be an allocation. There has to be an allocation formula to allocate that to individual wells. And there is an allocation formula, and they allocate it to the individual wells. No one is deceived by that. No one thinks that the Texas Railroad Commission, which they all know is where you're getting the information, reports it on an individual well basis. But it provides them some more information about how those wells compare. So there's an allocation formula. It doesn't mean that there's anybody that's being misled, that's being deceived about any of this. Then he says, well, you know, you didn't verify the data. You didn't verify the data. So what John Chisholm, the CEO, said after the fact, there was a glitch in the software, and he says, okay, we should have done a better job looking at it. What a responsible CEO would say, absolutely, it doesn't mean that you didn't do any verification or back checking. You've got a third-party provider who is this Ph.D. in this area, and there's no allegation here that he wasn't back checking and verifying. But in thousands of lines of code, you can have a problem. I used Google Maps to get from my hotel here to the courthouse this morning. Sometimes Google Maps sends me to the wrong place. I don't think they're trying to defraud me. I think they got the algorithm wrong. It happens in these large multi-line software programs where you can have a glitch in the algorithm, and it's a small thing that can have an effect. Was there any attempt to question further? When we talk about Dr. Collins and his expertise, is there no obligation whatsoever for the recipient of that information to question or, I hate to use the word, cross-examine or critique or peer review this kind of data? Or are they entitled to just accept it and trumpet it as successful? There's no particularized allegation that they didn't do that, that they didn't talk to him about the data, that they didn't ask him to stress test the data. There's no particularized allegation that that did not happen here. So what we've got on the record is he was the one. And, look, you could say, well, maybe they should have done more, especially as to these three that were put into this. He did not testify in this case. He's not even a defendant in this case. Pardon? He's not a defendant in the case. No, but he did not testify. He didn't testify. But you can look at the transcript, which is put in the evidence incorporated by reference by the plaintiffs in this case, which describes Dr. Collins and what he did and how he went about doing it. And then the third argument is, well, you're moving to a new algorithm. So Dr. Collins is moving to a new algorithm. But the company, and I'm just going to read to you in the record so that you have them, told the public from the beginning that the software would go through various versions. It said it's an internal application. The whole thing is just in its infancy and was being refined. That's at 107 and 112. Then it said Chisholm referred the app, this was in November 2014, as FRACMAX version 1.5. And then he said, and keep in mind that it's only been out, this is now later on in February of 2015 in the transcript, it's only been out three months with the latest version you're going to see, and that's at 1191. And then he says in 60 days we'll have version 3.0. You have different versions of software. Don't they say conclusively? Don't they use a pretty strong word, though? They do, and there's nothing that says that it isn't conclusive. You can have a software application. Remember, this is big data. This is big data analytics. So you're looking at 98,000 wells, and you're not looking at really one well and comparing it to another. You're looking at large amounts of wells. Does the evidence show that the defendants, before they publicly used this information, went back to Dr. Collins and said, look, we're getting ready to make a pitch to sell our product, and we want to use part of your, just a minuscule part of your information you gave us. We want you to recheck that and make sure it's right. We're just getting ready to make some strong representations based on that data. There's no evidence one way or the other in the record. But remember, these pages that had the incorrect data on them are three pages out of a 185-page presentation. They're just three pages out of this much, much larger presentation that's being made to investors. That means that the defendants passed over a whole lot of other stuff to get that seven-well comparison. And if they're going to bet the farm on that, I don't know whether it's scienter, but it tends to show scienter. Look, the case law is quite firm on this, that even if it was negligence or even gross negligence, which I don't believe it was either one of those, but even if it was, that's not sufficient. This has to be the severest departure from the standard of ordinary care for there to be a problem with this. And to take somebody who is as distinguished and reliable and who has a history of doing these types of applications and saying, you know, we can't rely on third-party providers when we're paying them good money to do this. What about the specific allegation that, specifically with regard to Mr. Chisholm, that the data was backchecked and validated? And there's no particularized facts indicating that it wasn't backchecked and validated by Mr. Collins. But they still missed the two or three pages out of the 180-something? Just like, I mean, I suspect that Google Maps backchecks and validates their data constantly, but still on occasion they will send you to the wrong location. Software can have glitches in it. It's not intentional. It's not even reckless. It happens in software because of the nature of the programming is so intense. And then when you test it, you run test codes through it, and if those test codes aren't handled exactly perfectly, then it creates a problem in the software. It doesn't mean anybody's doing anything to deceive. And there's no question here. You've heard the answers. No one's saying that the product doesn't work or that they're challenging the efficacy of the product or any of that. After the glitch was remedied, does the CF, whatever it is, CNF effectiveness show that the wells that were not misrepresented had less effective collection? No, they were still, they weren't as effective, but they were still, they still had an uplift in them, but they weren't as, the uplift, and that's all in the record, by the way. The uplift wasn't as great as it appeared from that glitch in the software. So there's nothing that says that the product doesn't work, and sort of the proof of the pudding in the oil industry is especially when you have a decline like we've had in this industry and everybody's saving every penny, they're continuing to buy the product. They're continuing to buy the product because the product works. Ultimately, the test of a product like this is does it work in a field that people think it enhances their production. So this case is not like Nathanson. I mean, the thread that is through all of the Fifth Circuit cases, all of these cases, the common thread in this, when the courts have found a strong and compelling inference of fraud, is that there is either a confidential witness, and there's not a single confidential witness in this case, and normally there would be if they had somebody, some person, somebody, a customer, a former employee, somebody who would say that they knew at the time that the information was wrong. That doesn't exist here. The other example is when they have a document. Nathanson is a good example of that. They had the patent itself that said that the drug application is not covered by the patent. They have an actual document. Here they don't have a document, and they don't have suspicious stock sales. This isn't a case where you're running up the stock price and then selling into the run-up. I mean, this would be a bizarre conspiracy where you're trying to mislead customers, or a bizarre scheme where you're trying to mislead customers about their very own data, which on the single-well leases is publicly available to them. It doesn't make any sense. You've got to, under Telabs, you've got to look at competing inferences, and the competing inference here is that it's a glitch, and that's what everything indicates that it is. The idea that this is a massive scheme does not work. Well, does Mr. Chisholm or any of the other defendants explain why these particular wells were selected to make the sales pitch with? I mean, these were all... I don't know if he actually explained that, but they were all within a small area of each other. They're all in the Eagleford. I mean, it could have been that. As the trial court said, it's because... It could have been. They looked at this and found these comparisons, maybe between the best and the worst, and they didn't go back and check with Dr. Collins to be sure about that before they started using it. They didn't see it, and they didn't expect it to be seen. I mean, this has been... It's not like this program was brand-spanking new at the time. It had been going on, and they had no reports from any customer, so they're showing this to customers. They're showing the customers their own data, and there's no allegation here that any customer is saying, hey, we've got our data wrong. That's not what our well's producing, our non-CNF well on a single-well lease. That's not what it's producing. Nobody is... The only one who raised any issue about this is a short seller whose, obviously, incentive is to drive the stock price down, who says that, look, these three wells look like they're wrong, and they are wrong. But it doesn't mean there's a massive problem here or a systemic problem within the organization, because that would have been discovered by somebody, by a customer, by, you know, the plaintiff's alleged that the SEC investigated this. No one has come out and said that there's a massive problem here. A special committee was formed. No one has said that, and none of that has come out in any of this. So... And the idea that you're evolving to a new algorithm, this is sort of the third prong of this argument, that they were evolving to a new algorithm. As I said, this was told to the shareholders from the beginning, and it is the storyboard of all software. You know, I've got a... I think I've got an Apple 8, you know, so it doesn't mean that the other ones were false and misleading applications, because you've gone through, you know, a series of them. And you're always looking to be more precise in how you provide the information. It doesn't mean that the information you provided before was wrong or that you knew about a glitch in the earlier software. And this is what Judge Bennett found. He said the fact that you're using algorithms doesn't mean that you knew that there was a glitch in the prior software. So let me just... And there's a comment about Judge Bennett that I thought might be worth mentioning, and it's, you know, where did he go wrong? He didn't go wrong at all. He didn't go wrong at all. I mean, Judge Bennett did everything that this court and the Supreme Court has asked him to do. He analyzed each one of the issues separately, and then he combined them and he analyzed them holistically. And he did it thoughtfully. He did it rigorously. He did exactly what a trial judge is supposed to do, and he came to the right conclusion. And so unless anyone has any additional questions for me, I will ask the court that you affirm the decision by Judge Bennett. Let's listen to this. Thank you, sir. Mr. Hubbachek, you have five minutes on the bottom. Thank you, Your Honor. Returning to our main points, which are, number one, that the data was being adjusted, and they said over and over and over again that it wasn't, and number two, that they said it was conclusive and yet they never verified it. Those are the main points of the claim. In addition, there's also the notion that they were changing the allocation algorithm during the class period and didn't tell anybody. Now, I understand that software evolves, and that doesn't always mean that there's something wrong, but here we have a situation where on the very day that Mr. Chisholm is trying to explain why the data went so wrong the day before, he says specifically that it was a problem. This is in paragraph 151 of the complaint. There was a problem with the allocation algorithm, and lo and behold, what was it that we were specifically working on for the past three months to change is the allocation algorithm. And what does he say? It's going to be more accurate now. So we're not talking about just because they happened to come up with a new version that that means that the old version must have been bad. Obviously my iPhone is just as good as my previous one, but in this specific instance, we're not talking about just an entire upgrade. If you look at this Court's decision in Abrams, there was a changeover of the internal controls, and in that particular case, they actually said in the beginning that the purpose of the change in the internal controls was to create more efficiency and to save money. There's no such allegation here. The inference that I think the most reasonable inference to draw from the fact that the very thing that went wrong is the very thing that they had spent three months changing, and it still wasn't ready, and all they could actually say was that we hope or we feel that it's going to be more accurate, I think that this Court can draw an inference that that means that they knew that there was something wrong with it for a significant period of time, several months plus, based on this Court's decision in Plotkin that says that you can infer previous conditions from present conditions. So that's number one. Number two, what's the thread that joins this case to other cases from this Court? It's Mr. Chisholm's status as one of the inventors of FRACMAX. We're not just saying that because he's the CEO that you should infer that he knows about this. He is an inventor. Now, the complaint alleges that he's an inventor, so all the comments about Dr. Collins can't change the fact that the most reasonable inference here is that Mr. Chisholm knew how this software application worked, and he knew that it did, in fact, adjust data, and that he knew that they didn't actually perform the verification. What did Chisholm invent? I'm sorry, Your Honor? What did Chisholm invent? He is an inventor of the FRACMAX application. That's it. That's the product. The FRACMAX application is the software application that's used to market the most important product. I thought he was an inventor of the stuff they put in the wells that made them perform better. He wasn't? No. We've alleged in the complaint that he's an inventor of the software application. Okay. And from that, I think this Court can draw inferences regarding his knowledge and regarding the other defendants' knowledge of its working, and you can conclude from that, based upon the information that was provided over a year's period of time, that they were severely reckless in making those false claims. I'd also like to just briefly touch on the notion that, you know, that FRACMAX is somehow in its infancy and that, you know, people wouldn't have really relied on it. I mean, the whole purpose of this marketing campaign was to get people to rely on it. Specifically, the target audience was new customers. They wanted to draw new customers in, so it wasn't in its infancy. They said ten times that it was a premier analytical tool. They said that it conclusively proved CNF's efficacy. They said that it was revolutionizing an industry and that it was the most compelling sales and validation tool that Mr. Chisholm had ever seen. The notion that this is, you know, in its infancy and that people should have their eyes wide open before relying on it simply doesn't make any sense in the context of the campaign that these defendants engaged in over a period of an entire year. And then lastly, I don't think that we've alleged in any way an irrational conspiracy, but I would like to suggest a motive to your honors. We put this in the opening brief. In the Telabs version on remand in the Seventh Circuit, Judge Posner talked about how an individual in an industry might make false statements in the hopes that the good news will overtake the bad news. What we have here is a situation where the defendants didn't want to say that maybe their evidence wasn't conclusive. They didn't want to admit that they actually did adjust data, and they were hoping that the changes that they were making to the allocation algorithm would, in fact, eliminate these problems. It just turned out that that gamble didn't pay off for them, as Judge Posner suggested. Thank you very much, your honors. Unless you have any questions, I'll submit. Thank you, sir. That concludes the oral arguments for today. This case will be taken under BAPA.