Oral argument — Howard Welgus v. Trinet Group, Inc.

Good morning. Good morning. May it please the Court, my name is Sean Williams. I represent Howard Welgus and the putative class. We're here on a dismissal of a securities class action that was brought under the Securities Act of 1933 and the Securities Exchange Act of 1934. I'd like to reserve three minutes for rebuttal, please. We'll try to help you there. Can I just ask you right off the bat, because I know there's a whole lot we could talk about. What's your best argument that the issue of falsity was adequately pled in the SAC? Well, the best argument that the issue of falsity was adequately pled is because this is one of those cases where the reasons why the statements were false are almost mirror images. For example, one of the key representations that we allege to be false and misleading is that risk management for this company was a core competency. In fact, it was a material weakness, and that's not... Is your best argument that this was a core competency issue, these people could not have failed to know about this? Is that what you're saying? No. I think that what we're saying is we're saying many things. But with respect to the specific claim or representation that this core competency was this company's core, I'm sorry. Oh, that was their... Okay. That risk management was the company's core competency. The fact that they didn't have employees that were trained adequately or with sufficient experience to do risk management was indeed evidence, a specific allegation, that core competency could not have been their risk management could not have been their core competency. And in fact, it was an admission of their own that they indicated was a material weakness. I'm sorry. Finish what you were saying. I was just saying that they admitted that in fact was a material weakness, that they didn't have the very employees necessary to do risk management. As I understand what the district court said, it said, and let me just sort of phrase this colloquially, these guys said back in whatever year, they were good at what they did. Now we have evidence later that they're not very good at what they did. But in order to make out a securities claim, you must show that at the time they made those statements, which seem like opinions to me, we're really good at management of risk. They had to believe that those statements were not true. And there's not much evidence here that shows what they knew or should have known at the time they made the statements. So could you tell me why I can, let's assume later that you've got evidence that they're not very good at risk management and they're not as competent as they said they were. What's the evidence that they must have known at the time they made the statements that they weren't accurate? Sure. The first point I want to address your question about are you suggesting that they may be opinion statements really quickly? Well, I think something like this is a core competency, we're really good at risk management, are not simply factual issues. In other words, really good at risk management says is a companion. I think they sound like opinions to me. They don't sound like pure facts. So if they are opinions, I guess my question is, is there any evidence that they knew that they were false when made? Sure. They had to know. And the reasons why are because the reasons they said they were very good at risk management, in fact, it being a core competency, because they analyzed every claim that came in the door and every client that came into the door, that they had superior access to data from insurance companies that simply wasn't available before, which allowed them to assess risk. The truth was that they did not analyze every claim that came in the door. In fact, they ignored the biggest swath of the clients and employees that had claims. With respect to the 66,000 employees that they acquired through the SOI business, they specifically admitted that they didn't look at them, they didn't look at every claim. In fact, when they did the due diligence on the entire business, they found that the reserves were not even up to industry standards. Can you go back to the statement that you think was misleading and just actually tell me the actual statement, not your characterization of it? Risk management is our core competency. There were more than just one statement. Risk management is our core competency. Okay. You were talking about analyzing claims. I'm sorry. Tell me what statement you think was false or misleading. There were six statements in the registration statement. One we've talked about a little bit. The next statement was we analyze every claim. Sorry. We analyze claims data for each claim, each client on an ongoing basis. Let me just see if you're thinking of a different statement than I am. I'm looking at one in a conference call that says we get state-of-the-art analytics on all of our claims. But that doesn't say we analyze every claim, does it? That doesn't say that we analyze every claim, but they were not getting state-of-the-art analytics. Okay. But the one you just said, that was the one I was focusing on because I couldn't find it. Where is the statement that we analyze every claim? It's in the registration statement. Yeah. Just identify it for me. The last statement in the registration statement that we allege to be false. We analyze claims data for each client on an ongoing basis. Above that, we assess all workers' compensation and medical health benefits risk on an individual client basis and annually adjust pricing to reflect their current risk. See, I read that statement, and it doesn't say we analyze every claim. It says we analyze each client and we analyze the risks of each client. But does it — I was trying to find a statement somewhere that said we analyze every claim. And I — You identified it in paragraph, I think you were looking at it, paragraph 108 and 109. But to the extent that they say that they analyze every client, that — Isn't analyzing a client different than analyzing a claim? Not for this company. No, it's not. Because every — every client, the only claims that could arise are claims based on the clients. So this company has what they call worksite employees. They have a couple hundred thousand of them. They make their money, they earn revenue on risk management for projecting to potential claims that these clients may submit. So if you are going to analyze every client, you must analyze every claim. And, in fact, what they told investors was because they had the data — You say that as a matter of logic, or you say that's what the amended complaint says? That's, in fact, what it says. And I just directed the court to — Well, but I just — I can't find a statement where they say we analyze every claim. Paragraph 109? Paragraph 109 of your — Sorry. Of your complaint. Of your complaint. Right. But tell me when they said we analyze every claim. They said, quote, that paragraph 108 is a quote from them in a conference call. Let me take a look. Okay. So if I can read it for you, it's the bottom of paragraph 108, page 42. We get state-of-the-art analytics on all of our claims. Correct. Every month we're able to get claims data so we can see how our clients are performing. Okay. I understand they made that statement. That was one we already identified. Can you identify a — and you may be right that that implies that they analyze every claim. Just a very specific question. Did they ever make a statement that we analyze every claim? They make a statement. I want to take that and look at the complaint and actually — Yeah, I have, too. I can't find that statement. Because the implication, your Honor, the implication from the quote that I just read to you is that they analyze claims data on every claim. And they, in fact, were not. But that's an implication. We're both looking for the actual language in High Fair but in the registration statement. And can you give us that, please? The registration statement says we analyze claims data for each client on an ongoing basis. And that means that they're analyzing each — Okay, but that doesn't say we look at each and every claim for every client, does it? That specific statement doesn't say that. But where does that statement appear in the registration statement? That statement, that specific statement is not in the registration statement. Okay, is it somewhere else that you referred to in the amended complaint? I believe that it is, your Honor. And I want to take — when I come back on rebuttal, I will address it very specifically for you. Very well. But the statements that the company made were that we have access to data that wasn't available prior to — That's the subsequent statement that you say shows the previous one was false, right? That's exactly. That's exactly right. And they simply didn't have it. They, in fact, admit that they weren't getting data on every claim. And, in fact, it was the — they were only even asking for data over $300,000 from the insurance companies when the problems that arose were for data that was — for claims that were under $300,000, for $50,000. Is that the gravamen of your complaint, that they said they were going to examine everything, every little thing, no exceptions, but, in fact, they didn't examine everything? They examined some things, but not everything. Therefore, they lied. Is that basically what you're saying? No, I think they gave the impression that because risk management was the best thing that they do, and the only thing that they do in which they make money, that because they were so good at it and that they could access and analyze every claim, or almost every claim and every client that came into the door, that the risks associated with unexpected claims was nearly nil. Okay, well, I mean, you're a smart guy. We'll take an algorithm. You don't have to analyze everything. You have to look at sequences. You look at patterns and so on. But that's not what they said. If they're so smart, they probably know how to use algorithms, and they can see how things are going, and they can determine whether something's going wrong. They don't have to look at every claim, do they, in order to — if they did — to fulfill what they say they're going to do. They didn't say that they used an algorithm to do so at all. No, I get it, but you're implying that they had to look at every claim. I'm simply pointing, as a matter of general knowledge, you don't have to look at every claim in order to do a really good job if you're doing your job. I wouldn't know the answer to that question because this company said or suggested that they look at every claim. And even if you don't look at every claim, let's just take, for instance, what they said about the 66,000 employees that they didn't look at at all. And so to the extent that they didn't look at them at all and ultimately found and admitted to investors that they should have done that, their due diligence, and, in fact, their due diligence didn't meet industry standards, ultimately, that resulted in significant losses, and they admitted that reserved for those claims. Yeah. And if this were a derivative suit, those would be great claims. My question is, why do they demonstrate that the statements when made were false? Well, it's not a derivative suit. That's right. And the impression that they gave to investors with respect to their ability to do risk management and assess claims at a level that no other company could do before, that the expectations of loss— Well, but now you're getting—be careful. You're going past what they said. They didn't say we can do this better than any other company could ever do before. You're now implying from stuff. My question is, you'll find out a year later or six months later that they're not very good at what they do. And so my question is, how does that lead me to know— why does that make it plausible that they were lying when they told you they were very good at what they do? Your Honor, I want to be sure that I have time for rebuttal. Think about it. Why don't you think about it and the answer to his question, and then come back and— I will. Very well. Now, this next part is tricky, as I recall, because you've got 15 minutes, but you've got three people. One of you has got 13 minutes, and the next two have one minute. You know that won't work. Yes, Your Honor. I'm speaking for all defense with respect to the common defenses. My co-counselors are speaking about the individual defenses that the underwriters and controllers— Well, if they're going to each—with respect, if they're each going to take a minute to speak about their individual defenses, I can't imagine that they're going to add anything to what we have in the briefs. My main concern was to make sure that Your Honors did not think that they were waiving them by not speaking. If we stipulated that in advance, would they be happy? I don't know that they would. Whatever you want to do, whatever you need to do to bill the clients as well. But the bottom line is that we just know from our experience that one minute—you're going to use up the time is what it boils down to. But we'll try. We'll give it a shot. So we'll tell you when the time is coming. This is a classic case of fraud by hindsight, that the plaintiffs are relying on optimistic statements that were made in the registration statement, comparing them to statements that were made later when things didn't turn out quite as rosily as hoped, and saying, aha, that shows that there was fraud. In this case, this Court's cases, the Supreme Court cases, make clear that that is not adequate. I want to talk about the specific arguments that my friend has made and why those are not supported, because as the Court has indicated, in many instances, he's taking liberties with what is said. He's making characterizations of what is said. And as this Court made clear in the Dearborn case by Your Honor, Judge Smith, the plaintiffs' conclusions are not facts. So we have to look at the facts. And in this case, there is no confidential informant, no confidential witness. There is no allegation of fact other than the words in black and white that were spoken by us and recorded in transcripts of interviews. Just to zero in on the statement that I have down in my notes, but I think you're going to tell us, is one of these kind of characterizations. But I have that you all represented that you received claims data for every client. Is that you're saying that that's not something you ever said? So we did say that. Let me clarify, though, because there is a gross confusion on the part of plaintiff with respect to who the client is. The client is the company, the co-employer. The WSEs are the employees who are co-employed by Trinet and the co-employer, the client. And that is reflected on SER 101, which is the principal statement of the registration statement from all of which these statements are drawn. Because it says that we take care. One of the ways in which we conduct risk management is by being very careful in the, quote, businesses we take on as clients. It's the businesses that are our clients. And so when we said, as we did in the quote Your Honor just pointed to, that we receive claims data for all of our clients, it's the client. It's the co-employer to whom we receive claims data. It does not mean we receive every little bit of data that might be available. Okay. I got it. Got it. Okay. So what were the true facts that came out on that point then? That you only receive claims data on claims above $300,000 or something like that? And, again, this is a mischaracterization of what was said, Your Honor. That's why I'm asking. Yes. So what were the true facts that came out? Of course. Because what was stated, and these statements about the $300,000 versus $50,000 limit are at SER 247 to 248. And what we said was historically we've asked our carriers for information on catastrophic claims over $300,000 that could potentially be in their pipeline. So we're asking for a different category of data than actual claims. This is for these big, potentially catastrophic claims, we want you to give us some foresight. What do you see coming down the road, right? And so now we ask them to tell us for any claim that they expect to be greater than $50,000 on a monthly basis so that we can track those through our pipeline. There's a potential claim for $50,000. Here's the diagnostic code. Here's the potential length of stay. So this will allow us to align the potential pipeline of large claims against our forecast. Counsel, this particular approach would be kind of boilerplate business school approach, would it not? It would indeed, Your Honor. And so what we're talking about is, of course, when you're dealing with actuarial forecasts, the anecdote is not what's valuable. I think this was the colloquy that you were having. It's the analysis of the data. It's the analytics. So don't deluge me with details every single bill and, you know, submitted by anybody. Give me the analytics. So when we're getting analytics for, you know, on the data, that doesn't suggest that we're looking at every little jot and tittle. And when we say that we get claims data for each client, it doesn't mean we get every little bit of data for every client. We get the claims data that is relevant. If you looked at every little bit, it probably would be worse in terms of analytics than the forest and trees problem, perhaps, right? Exactly, Your Honor. Exactly. It would be deluging with and making it impossible to see what was significant. Can I ask you about that specific point? There's a representation during a conference call in December of 2014 is that we get state-of-the-art analytics on all of our claims. Correct? That's right. And then don't – let's assume that subsequent facts show that maybe your analytics weren't state-of-the-art in 2014. Well, I don't think they do. I think one of the allegations that made is that we made representations that we were getting better data than was available to competitors three and four years before. But what you will not find in the complaint is any allegation of what data was available to those competitors three and four years before, by which you could compare them and whether we were receiving the same thing. Again, there's no allegation even about what we were actually receiving. So this is mere speculation that that was not accurate. So let me ask you about another statement, because this is one I don't understand, so you have to help me with it. At some point, this is about aggregate stop loss, and I'm not sure I understand the difference between aggregate stop loss and stop loss. You – there is a representation that we typically get stop loss on an aggregate basis. And then there's a later call in which somebody on your side says, well, it's not – aggregate stop loss isn't always cost-effective. So we don't – we don't – how do they square those two? Well, Your Honor, this is the most curious thing, I think, about the complaint. This is the allegation that the plaintiff makes by inserting brackets in what we actually said that says, but did not. This is the first instance I am ever aware of of fraud by bracket added by plaintiff. And – but that's not what we said. The question that was put in this dialogue, and this is at SER 248. 248. There was a – the question was, in a sense, what more could you do to cap the losses? Right. And Mr. Porter says, and I think this is a quote, we've also – we also can do aggregate stop loss, but again, when you look at it, the cost isn't really outweighing the benefit of doing that. Well, but the question was, what more could you do? Right. Okay. No. No, I understand. For example, one of the things in the question was, could you do stop loss on a regional basis? There was never any assertion that we did stop loss on a regional basis. No. But let me – let me focus on – I understand your point. Let me focus on what – what's concerning me. The disclosure says we typically include limits to our maximum aggregate exposure for claims at a given policy year. And Mr. Porter says we can do aggregate stop loss. We could in response to the question, but it – the cost really doesn't outweigh the benefit of doing it. Doesn't that suggest that the initial statement might not be accurate? No, Your Honor, and I want to point out, because I think this is very significant, that in the fiscal year 1510k, which – in which plaintiff says everything came clean finally and we saw the truth, the exact same language that Your Honor quoted about we typically include a term that limits our aggregate total loss and called stop loss, that's in the 1010k. Right. No. So I understand you've said it. You say repeatedly we typically do that. My question is why doesn't Mr. Porter's statement in 2015, which says – and maybe I don't understand what aggregate stop loss is, but it says we also can do aggregate stop loss, but when you look at that the cost isn't really outweighing the benefit of doing it, is that a statement that we aren't doing it or that we're doing it, but it's not cost effective? I can't – I don't know what that statement is. Well, Your Honor, I would think that they would have to allege something clearer that – to show that we weren't doing what we said we were doing. Well, but because it was in response to a question about – see, this is where I get back to your point. There was a response to a question about what more could you do. And then he says we could do this, but the cost does not weigh the benefit. Isn't that a statement that we're not doing it now? It's not a statement that we're not doing it. It's a statement that we're not going to do more of it. We're not going to do it on a regional basis, which we never said we were doing. We're not going to reduce the stop loss limits below that. And because there's a discussion immediately before that about whether you're going to reduce the workers' comp – or rather, the per claim limit from a million dollars to lower than that. At some point, if you're not taking on any of the risk, you don't get any of the benefit, is what he's saying. So we're not going to do more of this. And the very analyst, the plaintiff, cites as evidence of what was understood about this, and this is at – their quote is at ER 159, paragraph 121. It says that the company had showed no intent, quote, to further limit the likelihood of losses. So they understood it was that we weren't going to do more of this than we already were. And again, in the 10K at the end of 2015 – Yeah, I don't think there was ever a representation that total stop loss was being done. The question is whether any – whether aggregate stop loss was being done. Whether we were going to do more of that on a regional basis or other types that we weren't already doing. And at some point, if you're not taking on any risk, you don't get any of the benefit, and that's what Mr. Porter said. Isn't it correct that – I mean, the reality here, you're having a discussion of analytic tools available to a company that did what this company did. And sometimes you use one thing, sometimes you use another thing. It just depends what seems to be best under the circumstances. Something may work in one industry, some not in another. But the fact that somebody does or doesn't use a particular stop loss in a particular situation does not mean, does it, that the company is grossly negligent. It simply means that you've got a whole battery of tools. I'm going to use this one, I'm going to use that one, but I'm maybe not going to use this one today. You're right. What it adamantly does not say is what they impute to us by the brackets. It does not say that we don't do stop loss. Can I ask you to go back to the we get state-of-the-art analytics on all of our claims, that statement. So what's the difference between claims data and state-of-the-art analytics then? Because the first statement I asked you about, you said, oh, no, no, no, we just said we get stuff about the client. This statement obviously is focused on claims. So what's the difference? Well, the analytics is going to be an assessment. It's going to be, is your run rate higher than normal? Are you experiencing spikes in certain areas than other areas? I mean, I compare it to the information we get on a monthly basis in the law firm that tell us sort of like what practice groups are busy or not busy, but it's not telling me what every associate's time entries are. That would be too much information. That wouldn't be valuable to me. But the analytics that rolls it up and gives me some sense of what's going on, what the trends are, is this more or less than is expected in the same quarter and past years, that's analytics. It doesn't mean that I have to be looking at the specific detail of every claim as long as it's based on the claims information. I just want to note you've got 34 seconds before one of your partners over there wants to get up. And I want to go back to the risk management as a core competency. I mean, I think that it's even more than opinion. It's puffery and not actionable for that basis. But at the very most, it's an opinion, and they haven't established that we didn't believe it at the time. All they're able to point to is that at the end of the period, with respect to the internal controls for financial reporting, there were weaknesses. But with respect to internal controls for financial reporting in the registration statement, it is explicit that we had not done that, that our auditors had not looked at that for us, and if they had, they may well have found material weaknesses. And risk management, of course, means a lot more. It means 25 years of experience in the industry. It means being selective in the kind of clients we take on. We take on white-collar clients, primarily some blue and gray, but not the hard labor. We have risk mitigation procedures. So there is a lot in that that's not disproved by the fact that after these three blips, after we know these problems, we recognize those problems in saying we have some material weaknesses and we are addressing them by adding additional people to our staff, not that we had no staff to begin with. Thank you, Your Honor. Okay. Why don't the both of you come up? Because you've got a minute and 16 seconds, and you can each take turns, and you will have done your duty to your clients, we know. May it please the Court. I'm Alan Arfa from Paul Weiss for Defendant General Atlantic. And just to explain it, the principal issue here on the appeal is whether the plaintiff had sufficiently alleged primary liability by the issuer here, Trinet, and my colleague here has ably argued that issue, and we believe the plaintiff had not sufficiently alleged primary liability. If the Court finds that, you never reach the judgment. The only issue with General Atlantic is their sufficient allegation of controlled person liability under the two securities laws. There, if the Court, for whatever reason, finds the primary liability claims go forward, then you would address that. If, for whatever reason, the Court does address that, I just want to point out, the judge got the controlled person liability correct the first time. She did not get it correct the second time. The first time around, she correctly applied the standard in this circuit, which is actually pretty clear if you read the controlled person cases, that it's not enough just to be an officer or a director or be a substantial shareholder. It's not a status offense. On the other extreme, we don't require that you culpably participate in the wrongdoing. You have an intermediate standard that's very clear that requires some type of specific factual allegation showing either the person we're asking to be held liable is a controlled person. Involved in the day-to-day affairs or was involved with or controlled the preparation of the actual statements. Okay. You have used up all of your time and all of her time, but we're going to let you have some time. There's none of that in that. But you're going to keep going anyway. I just want to finish that sentence, that we're not involved. There were no allegations of either of those two factors. Okay. We'll give you a minute. Go ahead. Your Honor, it's fine. Laura Oswald on behalf of the underwriter. I note that with respect to the underwriter defendant, there is a separate and dispositive basis to dismiss the claims against Oswald. We were much like at the argument today, somewhat of an afterthought. We were adding to the complaint only on the third try, eight months after the claims were originally filed. The plaintiffs have done nothing in the lower court or in this court to explain how their complaint, on its face of the very first day, about the March 3rd disclosure and how that stunned the market. The quote from the complainant is, that stunned the market, and yet the plaintiffs claim that that was not sufficient to put them on trial and file a claim against the underwriters in the first instance of that year, and the second instance, and waited more than a year to add us. That's all. Okay. Thank you. Thank you for your argument. So you have some rebuttal time. Very quickly, looking at paragraph 111 with respect to the risk management issue. That's of the amended complaint, right? It is, Your Honor. And here's what they're asked. And before we move on off of the risk thing, the industry, I guess, is a bit littered with companies that have blown up in one way or another over the years because of poor risk management. Can you describe, like, how does the investor base get confidence that Trinet is different? And the answer is, the key to the risk, the risk of the business, is the data. The entire impression that was given to investors was because they had the data, the experience, the people, and the competency that the risks and the factors that caused the losses here simply would not happen. With respect to what analytics are as compared to data, data and analytics, when I looked it up, they were the same thing, and that's exactly what they intended to impress on the market. Analytics is what you do with data. I'm sorry? Analytics is what you do with data. That's right, but they weren't getting the data. They're not the same. And they weren't getting the data. They admit they weren't getting the data. And today they expect to explain to this Court why, what analytics mean. That's for the trial court. That's not what was argued in front of the district court. But did you plausibly allege that they weren't getting analytics? We plausibly allege that they weren't getting the data. In order to do analytics, you must get data. My question was, we get state-of-the-art analytics on every claim. Is there an allegation that they weren't getting analytics, or do you think that the fact that they weren't getting data on every claim? They weren't getting data, so they couldn't do the analytics on the data, which I think comes first. Well, they said they were getting it, and that's my question. Where were they getting it from, the analytics? What they said was that they were getting it from the insurance carriers, but later they said that they weren't getting it from the insurance carriers. And just lastly, Your Honors, there are two claims here, some under Section 11, some under 10b. And with respect to the Section 11 claims, it's only falsity and materiality. We've done that in spades, we believe. And with that, I will submit. Thank you very much. Thank you both. Thank you all. The case just argued is submitted, and the Court stands in recess for the day.