The Ninth Circuit just issued a loss causation opinion in In re BofI Securities Litigation, and it’s so beautiful, it gets so much right, it’s like staring at the sun, or the face of God. Birds sang, angels wept, my sinuses have been cleared, my freezer is defrosted, and all that’s left for me to do before I depart from this Earth is see Wonder Woman 1984 in theaters.

The background is a bit complex. BofI is the subject of two 10(b) securities class actions, covering different time periods.  The first alleges that the Bank lied about its lending practices, and the second alleges that the Bank lied about investigations into those lending practices.  Both cases were heard before the same district court judge, and the court dismissed both sets of claims on loss causation grounds, employing a particularly vigorous – nay, implausible – view of market efficiency. 

I blogged about second of those dismissals here, where I explained that the plaintiffs in that action had alleged that the fraud was revealed when a reporter for the New York Post filed a FOIA request and wrote an article about his findings.  The court rejected plaintiffs’ allegations of loss causation because anyone could have filed a FOIA request and obtained the relevant information; therefore, the court would assume that information obtained by the FOIA was already impounded in stock prices.

At the time, I called this “judicial magical thinking”:

Let’s take a moment to unpack the factual inferences here that the court is willing to draw at the 12(b)(6) stage: that unspecified investors made a FOIA request, that they got their response faster than the reporter’s own inquiry, and that they used that information to trade in sufficient quantities to completely offset the effects of the initial lie.  And that despite the fact that markets, apparently, can be expected to behave in this manner, these investors believed, ex ante, it would be cost-efficient to justify the time and expense of making the FOIA request in the first place so that they could exploit the information that the request – might! – reveal.

And, it should be noted, in drawing these inferences, the court remained untroubled by the fact that the stock did, in fact, drop upon publication of the Post article.

Forgive me if I have a little trouble accepting – without any additional evidence – that markets are imbued with this kind of near-mystical perfection. 

So I was not happy, in other words.

The district court pulled a similar move with respect to the other BofI class action.  There, the plaintiffs alleged the truth was revealed through a series of blog posts on Seeking Alpha, and through a whistleblower complaint filed by an ex employee named Erhart.  The district court reasoned that because the Seeking Alpha blog posts merely reanalyzed public data, they could not constitute corrective disclosures; further, the court held that because a whistleblower complaint only involves uncorroborated allegations, it, too, cannot reveal fraud and cause losses as a result.

Both cases were appealed to the Ninth Circuit, and this week, the Ninth Circuit issued its first opinion on the subject, holding that the whistleblower Erhart’s allegations were sufficient to establish loss causation. In so doing, the court used reasoning that, in my view, corrects the errors the district court made in both cases.

So, what’s good here?

First, the Ninth Circuit rejected the argument that the Seeking Alpha blog posts could not have revealed the fraud because they merely reanalyzed publicly available data.  Instead, the court held:

To rely on a corrective disclosure that is based on publicly available information, a plaintiff must plead with particularity facts plausibly explaining why the information was not yet reflected in the company’s stock price. The district court interpreted this requirement to mean that the shareholders had to allege facts explaining why “other market participants could not have done the same analysis and reached the same conclusion” as the authors of the blog posts. (Emphasis added.) We think that sets the bar too high. For pleading purposes, the shareholders needed to allege particular facts plausibly suggesting that other market participants had not done the same analysis, rather than “could not.” If other market participants had not done the same analysis, then it is plausible that the blog posts disclosed new information that the market had not yet incorporated into BofI’s stock price.

This reasoning implicitly rejects the district court’s holding in the case involving the NY Post article.  I.e., the mere fact that someone was able to do the analysis doesn’t create a presumption that someone else must have done the analysis earlier, let alone that these results were already reflected in the stock price.  Time is not a flat circle, things can happen at Time Two that haven’t already occurred at Time One.  The “modest” presumption that public information is reflected in market prices does not extend to all possible conclusions that could be drawn from that public information, let alone conclusions that require special effort, expertise, and investigation.

As for the whistleblower issue, the Ninth Circuit rejected the district court’s conclusion that an unconfirmed allegation cannot reveal fraud, and used particularly nice language in doing so:

To plead loss causation here, the shareholders did not have to establish that the allegations in Erhart’s lawsuit are in fact true. Falsity and loss causation are separate elements of a Rule 10b-5 claim. The shareholders adequately alleged that BofI’s misstatements were false through the allegations attributed to confidential witnesses. In analyzing loss causation, we therefore begin with the premise that BofI’s misstatements were false and ask whether the market at some point learned of their falsity—through whatever means. Viewed through that prism, the relevant question for loss causation purposes is whether the market reasonably perceived Erhart’s allegations as true and acted upon them accordingly. … If the market recalibrated BofI’s stock price on the assumption that Erhart’s allegations are true—and thus that BofI’s prior misstatements were false—then the drop in BofI’s stock price represented dissipation of inflation rather than a reaction to other non-fraud-related news.

This is a great quote; too often, courts allow concerns about the falsity of defendants’ statements to bleed into their analysis of the loss causation element.  I previously blogged about this problem, where I pointed out that courts (including, ahem, the Ninth Circuit in an earlier case) have sometimes held that mere allegations of fraud (or investigations, or what-have-you) cannot “reveal” the fraud to the market unless later events show those accusations to have been merited.  This does not make sense; it suggests a stock price drop occurring at Time One – upon publication of the allegations – is not a “real” loss unless there’s a corroboration of those allegations that emerges later at Time Two.  But the losses at Time One are the losses; shareholders will have experienced that pain regardless of whether subsequent revelations establish the reliability of the earlier disclosures.

Obviously, what’s going on with respect to these demands for corroboration at Time Two – long after the losses are felt – is that courts are suspicious that there was any falsity to be revealed in the first place.  Rather than say that, though, courts direct their skepticism toward loss causation, and demand especial proof that the corrective event really is revealing an underlying fraud.  But if a court is skeptical of the original falsity pleading, it should explain why the original falsity pleading was insufficient – and if the standard for pleading falsity was met, there shouldn’t be a second gamut where falsity is again tested with respect to the pleading of loss causation.

I further addressed this problem in my essay, Fact or Fiction: Flawed Approaches to Evaluating Market Behavior in Securities Litigation:

a number of courts have declared that announcements of investigations or lawsuits—whether instituted by the government, or internal inquiries—cannot, standing alone, cause losses for Section 10(b) purposes because an investigation or a complaint is merely an allegation rather than a confirmation of wrongdoing. This, of course, is a non sequitur; if a stock price represents traders’ view of the potential cash flows of the business, adjusted for risk, a credible possibility of fraud will cause them to reassess those risks and reprice the stock accordingly. Assuming the investigation was, in fact, caused by an underlying fraud—that is, if the plaintiffs are able to demonstrate the other elements of a Section 10(b) claim and show a causal chain between the investigation and the fraud they’ve alleged—there is no reason to treat stock price drops due to market distrust of the subject company as any less “caused” by the fraud as other kinds of price drops. An investigation is hardly an intervening event, and presumably, if there was fraud, and it is ultimately revealed—either with a full admission, or simply with disclosure of the underlying financial condition that it concealed—the stock will drop even further, to account for the fact that what was once an uncertainty has now become definite.

So I’m delighted with the Ninth Circuit’s analysis here.

I am, however, disappointed with Judge Lee, who dissented on this point.  Judge Lee put his concerns thusly:

But what if it turns out that Erhart’s allegations in his lawsuit are bunk? What if he is mistaken?

If that’s the case, per my Fact or Fiction piece, yes, it breaks the chain of causation between the underlying fraud and the public revelations, such that the losses were not proximately caused by the fraud.  Think of it this way: If the whistleblower is making it all up out of a dream he had, and by purest happenstance managed to identify a real fraud at the company, presumably that whistleblower would have lodged his allegations even if there had been no underlying fraud to be found.  If that is the case, the losses would have happened whether or not the firm was tainted, the losses were not caused by the fraud, and the case should be dismissed. QED.

But subsequent corroboration is not how we address that kind of causation problem.  Rather we ask — typically in discovery — whether there is reason to think the whistleblower allegations and the subsequent price drop were inspired by events entirely unrelated to the fraud.  At the pleading stage, there is no reason to think the whistleblower is an intervening cause; if a fraud is adequately alleged, and a whistle is blown by someone who appears to have knowledge of it, an inference is created that the two events are connected.  If later facts establish that not to be the case, fine, but the chain of causation hardly needs further examination on a motion to dismiss.

Anyway, that’s the good part of the Ninth Circuit’s opinion.  But every rose has its thorns, and, well, in this case, those appear in the Ninth Circuit’s analysis of the blog posts.  Though the court rejected the district court’s conclusion that analysis based on public information can never constitute a corrective disclosure, it also held that the blog posts here were not sufficiently credible for the market to have relied upon them, and thus they could not have revealed the fraud. As the Ninth Circuit put it:

The posts were authored by anonymous short-sellers who had a financial incentive to convince others to sell, and the posts included disclaimers from the authors stating that they made “no representation as to the accuracy or completeness of the information set forth in this article.” A reasonable investor reading these posts would likely have taken their contents with a healthy grain of salt.

Therefore, the shareholders have not plausibly alleged that any of the Seeking Alpha blog posts constituted a corrective disclosure.

That’s a helluva thing to conclude on the pleadings, especially since Joshua Mitts has demonstrated that traders do in fact rely on pseudonymous blog posts, and also make judgments about the posters’ credibility based on their track records.  Indeed, in this case, the plaintiffs alleged that the stock price dropped in reaction to the blog posts, which – for pleading purposes – suggests that traders did take them seriously, regardless of the Ninth Circuit’s post hoc assessments of what a reasonable investor would do.  It reminds me of that time the Ninth Circuit held that a statement was immaterial puffery despite the fact that the market moved in response to it.  See Police Ret. Sys. v. Intuitive Surgical, Inc., 759 F.3d 1051, 1060 (9th Cir. 2014).  Apparently, materiality judgments are for courts now, not traders.

Now, it’s true – as Judge Lee’s dissent notes, and as Joshua Mitts has documented – we have a short-trader market manipulation problem.  Sometimes, pseudonymous bloggers post false accusations in order to cash in on the resulting price drop.  But that issue is hardly solved by holding that analyses which really do identify underlying misconduct (which, again, we’re assuming for pleading purposes) cannot constitute corrective disclosures for the shareholders who bought at an earlier time point.  Indeed, the Ninth Circuit’s reasoning leads to the perverse conclusion that the blog posts of the bad actor manipulative short-sellers – the ones who manufacture accusations in order to force a stock price drop – are also immaterial, have no effect on stock prices, have caused no harm, and therefore the posters are immune from punishment.  That is the opposite of what we should be doing.

Well.  Nobody’s perfect.

BLPB Readers,

Exciting news!  On the morning of October 14, 2020, the Systemic Risk Council is offering a webinar on Ensuring Financial Stability: Relaunching the Reform Debate After Pandemic Dislocation.  The Agenda looks fantastic! A brief summary of the program is below:

The stimulus response to the global pandemic has surfaced new debates and highlighted lingering questions about the role of central banks, accountability, reform, and the roles of levered markets and shadow banking.  This Systemic Risk Council program brings together leading voices to explore how the financial industry, regulators, and policy makers can address key issues around bank stability, resolution, and the mounting leverage in the global economic system.

I usually leave the arcana of contract interpretation principles to the specialists, but every now and then I apparently dip my toe in, and this is another of those weeks.

VC Laster’s opinion in In re Anthem-Cigna Merger Litigation tells a truly wild tale.  In brief, Anthem and Cigna agreed to merge, but Cigna’s CEO, David Cordani, wanted to helm the combined entity. When it became clear he wouldn’t get the CEO spot, he began a campaign of sabotage, assisted by Cigna executives who supported his ambitions.  Thus, Cigna’s top leadership hired a PR firm to trash talk the merger while concealing the firm’s involvement from Cigna’s Board.  At one point, Cigna’s General Counsel personally leaked certain letters so they would be disclosed publicly, then tasked her head of litigation with conducting an internal investigation to discover the source of the leak.  Cigna’s foot-dragging with respect to integration planning and responding to regulators was so profoundly passive-aggressive that I felt my blood pressure rising – and this is despite the fact that Anthem was not exactly an angel: it lied in federal court proceedings, and the overall merger would certainly have reduced competition in an already-consolidated industry.

In any event, both sides sued, and Laster ultimately concluded that while Cigna’s breaches were many and varied, the merger would have failed anyway due to regulatory opposition.

But entertainment-value aside, a technical point caught my attention, and that’s the burden of proof framework that Laster adopted, and which is in some tension with the framework articulated by the Delaware Supreme Court in Williams v. ETE, 159 A.3d 264 (Del. 2017).  Notably, however, it’s not clear that the framework was outcome-determinative, and even if it was, the party harmed – Anthem – endorsed Laster’s interpretation in its post-trial briefing, so this couldn’t be grounds for an appeal.

The issue: Sometimes, a party’s obligation to perform on a contract may only be triggered if a particular condition occurs. In that case, if the condition does not occur, the obligation is not triggered.  But, if the party prevented the condition from occurring by breaching its own contractual responsibilities, the party is obliged to perform.

In this case, the performance obligation was the merger itself, subject to the condition that there be no regulatory injunction preventing consummation.  Such an injunction did issue, however, when federal courts held the merger would be anticompetitive.  Thus, the parties were nominally relieved of the obligation to close.  Anthem, however, claimed that Cigna’s lack of cooperation breached its obligations under the merger agreement, which was the cause of the regulatory failure.

Laster agreed that Cigna was flagrantly – nay, ostentatiously – in breach of its obligations to cooperate with regulators and to use its best efforts to close.  Thus, the only question was whether those breaches were responsible for the fatal injunction.  And here’s where the burden of proof issue comes up.

In Williams, the Delaware Supreme Court held, “once a breach of a covenant is established, the burden is on the breaching party to show that the breach did not materially contribute to the failure of the transaction.”  Words to that effect are repeated a couple of times in the opinion (e.g., “if a proper analysis of ETE’s covenants led to a conclusion that ETE breached those covenants, the burden would shift to ETE to prove that its breaches did not materially contribute to the failure of the closing condition”), and CJ Strine in dissent also wrote, “As the Majority notes, under our law if a party establishes a breach of a covenant to bring about a condition at closing, the burden is on the breaching party to show that the breach did not materially contribute to the failure of that closing condition.”

So these statements would suggest that all Anthem needed to do was show that Cigna breached its obligations.  After that, the burden would be on Cigna to show that the breaches did not materially contribute to the failure of the condition, i.e., were not responsible for the issuance of the injunction.

But Laster didn’t see it that way.  Both the Williams majority and the dissent cited in support Restatement § 245 comment b.  That section of the Restatement says:

Where a party’s breach by non-performance contributes materially to the non-occurrence of a condition of one of his duties, the non-occurrence is excused.

And the comment reads:

Although it is implicit in the rule that the condition has not occurred, it is not necessary to show that it would have occurred but for the lack of cooperation. It is only required that the breach have contributed materially to the non-occurrence. Nevertheless, if it can be shown that the condition would not have occurred regardless of the lack of cooperation, the failure of performance did not contribute materially to its non-occurrence and the rule does not apply. The burden of showing this is properly thrown on the party in breach.

To Laster, this suggested a more complex burden shifting framework.  First, Anthem would have to show that Cigna breached its obligations in a manner that contributed materially to the issuance of the injunction.  After establishing that, Cigna would have the opportunity to offer an affirmative defense that even if it had cooperated, the injunction still would have issued.  Laster found Williams unclear on this point, I believe, because despite the language I quoted above, it did not distinguish between proof that the breach contributed to the condition’s failure, and proof that lack of a breach would still have led to failure – inquiries that he believed were distinct. And there is precedent from other jurisdictions supporting Laster’s reading, for example, Cox v. Snap, 859 F.3d 304 (4th Cir. 2017) and Cogswell v. CitiFinancial Mortgage, 624 F.3d 395 (7th Cir. 2010).

And so, that’s how he proceeded.  Under this framework, Anthem was unable to make the initial showing that certain of Cigna’s breaches contributed to the injunction, but it did succeed in making that showing with respect to other breaches.  And at that point, the burden switched to Cigna which, Laster concluded, was able to establish that the outcome would have been the same regardless.

Now, all this might be angels-on-pinheads – after all, there were only two Delaware cases that anyone cited regarding this matter, Williams and WaveDivision Holdings, LLC v. Millennium Digital Media Systems, L.L.C., 2010 WL 3706624 (Del. Ch. Sept. 17, 2010) – so it apparently doesn’t come up all that often.  But it does come up.  The Restatement itself offers an illustration involving earn-outs (and, umm, also one involving a husband murdering his wife and then himself, which … okay, we’ll save that for a critical gender studies post), and those apparently have become more popular due to covid-related uncertainty (earnouts, I mean, not murder-suicides), so this is an issue that could have a subtle effect on outcomes in future cases.