Saturday, September 3, 2016
At this point, it almost feels like I’ve been following the securities fraud case against Halliburton my entire career. I was grimly amused when I heard that the Fifth Circuit had granted an interlocutory appeal – again! – to hear a challenge to class certification – again! and I diligently tracked the docket on Bloomberg, only for it to belatedly sink in that – wait a minute, I actually live in New Orleans now. I can go see the oral argument in person, live, in color.
So I did.
It was … interesting.
[More under the jump]
I’ve discussed the Halliburton case in numerous previous posts, but the basic issue is how courts must address the question of stock price impact at the class certification stage of a Section 10(b) action.
Where we left off, the Supreme Court held in Halliburton Co. v. Erica P. John Fund, Inc., 134 S. Ct. 2398 (2014) (“Halliburton II”) that plaintiffs are entitled to a presumption that material misstatements in an efficient market affect stock prices, but that defendants may attempt to rebut that presumption at class certification. How, exactly, defendants are supposed to accomplish this is left to the mists. Previously, in Erica P. John Fund, Inc. v. Halliburton Co., 563 U.S. 804 (2011) (“Halliburton I”), the Supreme Court held that loss causation (i.e., the stock price drop upon disclosure of the truth) cannot be considered at class certification (because a stock price drop at the end of a fraud period is not the same thing as the price impact of the lie earlier in the period), and in Amgen Inc. v. Conn. Ret. Plans & Trust Funds, 133 S. Ct. 1184 (2013), it held that materiality cannot be considered at class certification (because the existence of materiality – a prerequisite for price impact – is a requisite element of all Section 10(b) actions and thus common to all investors).
Defendants in multiple cases have tried to persuade courts that if they can both show that the stock price did not rise at the time of the initial misstatement, and that the stock price did not drop when the truth was disclosed, that is enough to rebut the presumption that the misstatement impacted prices initially
The problem with that theory, as I’ve previously argued, is that there is almost never a price reaction to the initial misstatement, because most misstatements serve to maintain stock prices, rather than inflate them. And there is always some kind of disclosure, and a stock price drop, at the end of the class period, because why else would plaintiffs even file a case in the first place?
So as a practical matter, the defendants’ argument – that there must either be a price rise to start, or a price fall in reaction to truth later – means that ultimately, disputes will often center on whether the disclosure that prompted the later drop was, in fact, corrective of the earlier disclosure, or whether it was simply a report of unrelated bad news. If it was not corrective, according to the defendants, the presumption of price impact of the initial lie is rebutted.
This is, of course, logically flawed, for the reasons identified in Halliburton I. Even if defendants are correct that the later disclosure did not correct an earlier misstatement, all that means is that the truth was never revealed. But that doesn’t tell us whether the initial misstatement impacted prices: the stock price may still be inflated, even today; or (more likely) the fraudulent information slowly became stale and the stock price gradually drifted down in undetectable ways. All of this is a serious problem for plaintiffs when they try to prove the element of loss causation. But it doesn’t show that the initial misstatement did not impact prices in the first instance. If plaintiffs are entitled to a presumption of price impact – based on the materiality of the misstatement (off limits at class certification) and the efficiency of the market – the argument that the truth never clearly came to light is simply orthogonal to the point.
Nonetheless, this is the argument on which the defendants hang their hat in Halliburton: the disclosure at the end of the class period did not, in fact, reveal any earlier fraud; therefore, the plaintiffs have no basis for presuming there was price impact at the beginning of the class period. The district court refused to consider this argument; it held that battles over whether a particular disclosure was, or was not, corrective were inappropriate for class certification, and certified the class. The Fifth Circuit granted the defendants' petition for an interlocutory appeal.
Which brings us to the oral argument before the Fifth Circuit, which was held on August 31 (you can download the audio in .mp3 form here).
In keeping with their strategy in the district court, plaintiffs continued to leave essentially unchallenged defendants’ central premise: that if there was no stock price drop in reaction to the truth, and no price rise in reaction to the initial disclosure, price impact has been disproven. And because in Halliburton – as in most securities cases – the stock price did not rise in reaction to the initial misstatement, plaintiffs began from a position of weakness.
I’m not certain why plaintiffs have chosen to give up so much ground. I wonder if they’re assuming that Greenberg v. Crossroads Sys., 364 F.3d 657 (5th Cir. 2004) is still good law; in that case, the Fifth Circuit did hold that plaintiffs must prove price impact by showing that that the stock rose in reaction to the misstatement, or fell upon disclosure of the truth. However, I believe the logic of Greenberg – if not its actual holding – has been undermined by Halliburton I.
In any event, the plaintiffs in Halliburton do not seem to be contesting the basic proposition that whichever party bears the burden, if the stock price did not move when the misstatement was made, and if there is no evidence of a corrective disclosure followed by a price drop later, the misstatement cannot have affected stock prices. Instead, the plaintiffs argued, via David Boies, that whatever other evidence may be considered on these points, it cannot include whether the disclosure was, in fact corrective. He offered little to the panel by way of what evidence can be considered – which is a critical problem, once you've conceded that the defendants do have the right to put on some evidence of something, and that price impact means either movement at the beginning of the class period, or the end.
Unsurprisingly, the panel appeared unconvinced. Judge Elrod focused solely on the question whether remand, or outright reversal, was in order (never a good sign for the appellee), while Judge Higginson searched for any hint in the record that the district court actually considered evidence that the disclosure was, in fact, corrective, thus sparing the court the need to decide this issue. Higginson even stated, repeatedly, that in Halliburton II, the Supreme Court relaxed Amgen’s earlier holding that questions common to the class need not be entertained at class certification; in his view, Halliburton II allows courts to consider questions common to the class at the certification stage, so long as the defendant bears the burden.
I almost cried when Higginson asked both parties’ counsel whether any academics have weighed in on interpretations of Halliburton II, and neither side could name anything. To be honest, I’m not aware of any law review articles on this precise point either, because academics have mostly been (1) messing around with blog posts (sidebar: I guess now we know what kind of impact blog posts have on judges), or (2) have already discussed the issue, ad nauseam, in articles regarding causation in securities cases, and didn’t feel Halliburton II changed the landscape much.
I did think Boies’s most powerful argument was that this question of “fit” – whether the disclosure really was corrective – may be litigated as a loss causation argument at the motion to dismiss, and at summary judgment. He made the critical point that the district court in this case is currently considering that very question, in a pending motion for summary judgment. The issue of fit will be decided in this case, he made clear – the only question is whether it needs to be decided at this stage.
But he didn’t tie that argument to the broader point: Fit doesn’t tell us whether the stock price was inflated initially.
(To be fair, there could be some reason to litigate fit, but not under these circumstances. Defendants could argue that a disclosure in the middle of the fraud period was corrective, and that the stock price did not react, and this fact implies that the earlier lie had not impacted prices either - leaving plaintiffs in the position of arguing that price didn't react because the disclosure was not corrective. This, not coincidentally, is similar to the argument currently under consideration before the Second Circuit in the Goldman Sachs case. But it's the opposite of the argument in Halliburton, which starts from the premise that if a fraud was never exposed in the forest, it never made a sound.)
So, where things are left: I hate predicting outcomes because I am very very bad at it, but my read would be that the Fifth Circuit will hold that if defendants show both that there was no stock price rise at the start of the class period, and no price reaction to a corrective disclosure at the end of the class period, price impact has been rebutted. And they will further hold that district courts must consider at class certification whether a purported corrective disclosure was, in fact, corrective – suggesting that if the truth was never revealed, plaintiffs cannot get a class certified.
Such a holding would, of course, put us in nearly exactly the position we were when this case began. See Archdiocese of Milwaukee Supporting Fund, Inc. v. Halliburton Co, 597 F.3d 330 (5th Cir. 2010). So, six-odd years of litigation, two Supreme Court trips, and ....