This week, we got two denials of class certification in 10b-5 securities cases involving meme stocks. The first concerned Bed Bath and Beyond, the second concerned a fintech called Rocket Companies, which is not one of your more famous meme stocks, but apparently met the definition for 2 days out of a 2-and-a-half month class period. One case presented a refreshingly accurate application of current doctrine. The other presented a clarifying illustration of the doctrinal mess created by the Supreme Court’s decision in Goldman Sachs v. Arkansas Teacher Retirement System and its subsequent interpretation by the Second Circuit.
Section 10(b), and Rule 10b-5, prohibit fraud in connection with securities transactions. Among other things, they prohibit corporate executives from publicly lying about a company, which typically causes the stock price to go up – only to crash again when the truth is revealed.
But when a plaintiff tries to sue in these cases, she confronts a fundamental problem: Fraud claims require proof of reliance. And most stock purchasers may have trouble proving they relied on any specific false statement. Maybe the investor didn’t hear the statement personally; maybe they relied on analyst advice – or an index. Plus, most investor losses are too small to be worth suing over individually; the claims only make sense brought as a class action. But if each investor has to prove that she personally relied on the false statement, there are too many issues to adjudicate collectively.
The solution to this problem is the fraud on the market doctrine, adopted by the Supreme Court in Basic, Inc. v. Levinson (1988), and reaffirmed in Halliburton Co. v. Erica P. John Fund, Inc. (2014). The doctrine consists of two presumptions – I call them the “objective” presumption and the “subjective” presumption – that benefit 10b-5 plaintiffs to help them satisfy the element of reliance on a classwide basis.
The first presumption is that, in an open and developed market – wide trading, lots of analyst coverage, and so forth – material information impacts stock prices. The second presumption is that investors, subjectively, rely on prices as an unbiased assessment of market value when making an investment. Together, the presumptions establish that the defendants’ fraud impacted stock prices, and investors relied on those prices – so, by syllogism, investors relied on the fraud. Therefore, plaintiffs can satisfy the element of reliance at trial and – crucially – because these presumptions apply to all purchasers, plaintiffs can also get a class certified.
But, as Halliburton made clear, these are only presumptions; defendants are permitted to try to rebut them, and if they do so successfully, they can prevent class certification.
Which brings us to In re Bed Bath & Beyond Securities Litigation, 2024 WL 4332616 (D.D.C. Sept. 27, 2024).
Ryan Cohen took a 10% stake in Bed Bath & Beyond. Later, as the company ran into trouble, he issued a tweet that included a “moon” emoji, which retail investors took to mean he planned to stay in for the long haul. Other Cohen filings – a 13D, a Form 144 – also suggested he’d stand pat. Retail investors piled in, the stock price rose, but tanked when it was revealed Cohen sold his stake. A class action followed, alleging Cohen misrepresented his intentions. The court denied Cohen’s motion to dismiss, leading to the motion for class certification.
In evaluating the motion, the court began by correctly identifying the objective and the subjective presumptions that comprise the fraud on the market doctrine. The court then observed that during this period, Bed Bath & Beyond’s stock was going nuts. It was entirely untethered from fundamentals because it was a meme stock. Under those circumstances, Judge McFadden refused to presume that Cohen’s false statement affected its price.
But think about this for a moment. The stock was erratic; you can’t simply presume any particular piece of information is affecting prices. Yet, plaintiffs might still be able to affirmatively prove the statement affected prices. If so, they’ve satisfied the first half of the fraud on the market syllogism – the stock price was affected. So long as the second half remains intact – a presumption that investors relied on stock prices – they can still prove their case classwide.
In fact, that’s precisely the argument I laid out before in my blog post about a case against Robinhood, In re January 2021 Short Squeeze Litigation.
Judge McFadden understood that, as well, because he did not rest on his conclusion that there could be no presumption of price impact. Though he didn’t organize his reasoning the way I’d recommend, he still recognized that plaintiffs might simply introduce direct evidence that the price was impacted – but they had failed to do so. As he put it:
Bratya argues that price impact from the August 12 tweet is visible to the naked eye. Pl.’s Reply at 17. A graph of BBBY’s price that day shows the price steadily rising from the market opening until 10:41 am, when Cohen sent out his tweet. Fischel Report ¶ 40. And it continues to rise at a steady rate before rising more dramatically around 12:30 pm. Id. Fischel claims that this steady, continuous rise does not register any price bump following Cohen’s tweet. Id. ¶ 40. To the Court’s eyes, however, the graph in fact does show a bump in BBBY’s price in the hour following the tweet, with a more dramatic price increase occurring just two hours later, between 12:30 pm and 1:15 pm. But neither Fischel nor Cain offer analysis on whether these blips are significant or noise. And having been lectured by both parties on the importance of statistical rigor, the Court will not make any conclusions about price impact by eyeballing one graph.
Thus, since there could be no presumption of price impact, and there was not sufficient affirmative direct evidence of price impact, the first leg of the fraud on the market syllogism failed; so, no classwide proof of reliance, and no class certification. I’m not going to weigh in on whether McFadden correctly evaluated the weight and direction of various pieces of evidence, but I admire his clarity in identifying the correct questions: namely, was the market of a character such that we can presume price impact and, if not, is there other evidence of such impact?
Though McFadden did not, we could go further. Remember, the second half of the syllogism is a presumption that investors rely on stock prices. What does that even mean? To be honest, it’s doctrinally unclear, see Donald C. Langevoort, Basic at Twenty: Rethinking Fraud on the Market, 2009 Wis. L. Rev. 151, but it does at minimum suggest investors believed the price to be validly set by supply and demand through honest investor assessment of information available. And it’s very difficult to say this condition would be met for a meme stock. For a meme stock, the entire point is that retail investors collectively decide to manipulate the stock price to something other than what a sober assessment of the company’s prospects would suggest. At bare minimum, we might say that even if some retail investors “relied” on stock prices, so many of them did not that we’d create individualized issues for defendants’ rebuttal rights.
The upshot being, it may very well be the case that individual investors relied on Cohen’s statements – and if they brought claims individually, they could prove their reliance. But it’s more difficult to presume reliance on a classwide basis.
But then we turn to the second meme-stock-class-cert case, Shupe v. Rocket Cos., 2024 U.S. Dist. LEXIS 178076 (E.D. Mich. Sept. 30, 2024).
Rocket Companies is a fintech that allegedly lied when it claimed that rising interest rates were having no impact on demand for its loans. When the truth of declining margins and loan volume was disclosed, its stock price dropped. Shareholders brought a 10b-5 class action, and the court denied class certification on the ground that the defendants had sufficiently rebutted the fraud on the market presumption.
What evidence did they offer?
As with Bed Bath & Beyond, the defendants focused on the objective presumption – namely, the presumption that their false statements impacted market prices. To rebut that presumption, they first brought in an expert who pointed to public information about how rising interest rates were likely to harm Rocket’s business, including the risk warnings included in Rocket’s own SEC filings. She also demonstrated that market analysts at the time were focused on these indicators, and did not appear to rely on Rocket’s statements to the contrary.
In securities litigation parlance, I find this evidence similar to what we usually call “truth on the market” evidence, namely, evidence that sufficient truthful information was swirling around to offset the impact of the defendant’s lies. Except “truth on the market” evidence is exactly the analysis that the Supreme Court held in Amgen Inc. v. Connecticut Retirement Plans, 568 U.S. 455 (2013), may not be considered on a motion for class certification. And in Goldman Sachs v. Arkansas Teacher Retirement System, the Supreme Court reaffirmed the correctness of the Amgen holding.
Now, the Goldman Sachs decision by the Supreme Court, echoed by the Second Circuit on remand, also suggested that if the original false statement is much more “generic” than the revelation of the truth, there may be an inference that any price drops associated with the disclosure did not reflect the dissipation of artificial inflation introduced by the lie.
Is that this case?
Yes, according to the Rocket court. The false statements were that Rocket saw “strong consumer demand,” and that its networks were “growing.” The corrective disclosure was that Defendants expected “[c]losed loan volume of between $82.5 billion and 87.5 billion”; (2) “[n]et rate lock volume between $81.5 billion and $88.5 billion”; and (3) “[g]ain on sale margins of 2.65% to 2.95%,” – numbers that were far lower than Rocket had reported previously.
See? The defendants told everyone things were going well, and then disclosed specific figures showing they were going poorly, and the stock price tanked. How could anyone possibly think the two were connected?
(That was my sarcastic voice.)
This is exactly what I predicted would happen after the Second Circuit’s Goldman decision: namely, plaintiffs and defendants would fight about whether corrective disclosures are sufficiently similar to the initial lie, untethered any determinate standard other than the trial judge’s gut feeling.
Does anyone remember when, in the context of loss causation – which, nominally, courts may not consider on class certification – courts held that disclosures need not be “mirror images” of the initial lie, because that would functionally require corporate managers to admit to fraud, see Freudenberg v. E*Trade Financial Corp., 712 F. Supp. 2d 171 (S.D.N.Y. 2010), and “would eliminate the possibility of 10b–5 claims altogether” In re Williams Sec. Litig., 558 F.3d 1130 (10th Cir. 2009)?
And remember when the Second Circuit held in Goldman that “the question here—whether there is a basis to infer that the back-end price equals front-end inflation—is a different question than loss causation, and, in light of Goldman, requires a closer fit (even if not precise) between the front- and back-end statements….” Op. at 54 n.11?
Yeah, I remember that, too.
And another thing: New Shareholder Primacy podcast up! This time, Mike Levin and I talk 10(b) actions by shareholders of pre-merger SPACs, and the Tesla pay case (not that one – the other one). Here at Apple; here at Spotify; here at Youtube.