It is easy to armchair quarterback a Supreme Court argument, but there are a number of issues I wished had been more fleshed out.
The difference between proving loss causation and proving fraud-on-the-market reliance. The justices seemed to look for a way to distinguish requiring plaintiffs to prove that a market for a particular security was efficient from requiring plaintiffs to prove loss causation. As Professor Black and I both noted earlier, even Halliburton’s counsel conceded that the Fifth Circuit went way past Basic in requiring that plaintiffs prove loss causation as a condition to the fraud-on-the-market presumption of reliance.
The Supreme Court has been clear in Dura and more recently this term in Matrixx that reliance and loss causation are separate issues. Again, I don’t find Halliburton’s efforts to distinguish “price impact” from loss causation to be persuasive. I doubt the Court will want to want to re-muddle the issue by conflating proof of “price impact” with proof of reliance.
To be clear, plaintiffs still have to prove that a market is efficient in order to get the fraud-on-the-market presumption of reliance. What does this proof look like? It ranges from event studies to expert testimony that show that prices in a market for particular security typically change in response to new public information. Trial courts have managed to sift through this kind of proof for two decades.
Halliburton’s attorneys are now recasting their litigation at trial and the two lower court opinions to say that they rebutted this presumption of an efficient market by asserting that prices did not respond to particular corrective disclosures.
Aside from the time travel issue of trying to retry a case on appeal, what’s the problem with that? Well for one thing, requiring plaintiffs to prove that a particular disclosure – whether a misleading disclosure or a corrective one – caused a particular price movement would undermine the distinction between proving loss causation and reliance.
There is a larger practical problem for investors: mixed disclosures. Even if plaintiffs can prove that market prices for a company’s securities typically react to new information, it is very difficult for plaintiffs to prove that a particular statement had a particular effect on price. Companies tend to make statements in bunches making it difficult to distill the effects of the first false statement on the stock price.
Let’s assume Gerding Enterprises is publicly traded and I am the chief executive officer. A particular false material statement (“Professor Gerding was born in Texas”) may be bundled with other statements false (“Professor Gerding robbed a bank yesterday”) and true (“Professor Gerding lives in New Mexico”). (I’ll leave it to the reader to determine whether the lie of being born in Texas would increase or decrease the price of my stock.)
Similarly, corrective disclosure may be made in bunches – with corrective truths mixed with other statements either true or false. “Professor Gerding was in fact born in New Jersey, not far from the birthplace of Justices Scalia and Alito. He won the 2011 Nobel Prize for Literature. He was convicted yesterday of arson.” Correcting the misstatement of my birthplace may cause the stock price of Gerding Enterprises to move back to where it would have been had the lie not been made. But the fresh lie about my non-existent Nobel might inflate the price, and the false news of my criminal conviction would depress the price. If the price of my stock doesn’t move, doesn't move enough, or moves in an odd direction after this second bundle of disclosure, does that mean that the market for Gerding Enterprises stock was not efficient with respect to the lie about my birthplace? Or does it mean there is a mess with untangling the effects of all the statements made.
Although the particular details I picked are fanciful, the problem of mixed and bundled disclosure is definitely not. The corrective disclosure in Halliburton itself involved "multiple pieces of negative news." In Oscar, the issuer (a telecom company) made a corrective disclosure that it had overstated the number of telephone lines it installed at the same time it disclosed that it missed analyst estimates on earnings. The Fifth Circuit rejected class certification in Oscar because the plaintiffs could not isolate the effects of the corrective disclosure from the other statements. But might lower earnings have some relationship to non-existent revenue from non-existent telephone lines?
Ultimately, Plaintiffs still have to untangle the causal effects of all the statements and prove that a particular misleading statement caused their loss. But, again, the issue is when. Not only is this causation issue common to all plaintiffs, but if investors have to prove it at class certification, they will have minimal discovery. The assertion by Halliburton’s counsel that trial judges have flexibility to allow for intensive discovery at the class certification stage did not sit well with Justice Scalia or Chief Justice Roberts.
Bad incentives. Moreover, companies can seek to obfuscate and frustrate plaintiffs from ever getting a class certified by mixing in extraneous true and false statements when they make either an initial misleading statement or a corrective disclosure. Professors Fisch and Spindler and Fisch, among others, have made this point persuasively. Disclosing the telephone book is the best way to defeat securities litigation if not to cloak outright fraud. Again, it would be hard to prove this kind of strategic obfuscation with minimal discovery.
Justice Sotomayor asked: [d]oesn’t a lack of response to a disclosure – couldn’t it be in some situations reflective of an inefficient market? It could, or it could have other explanations including that the market is indeed efficient but that either
• the market already learned the truth (the truth-on-the-market defense a la Judge Easterbrook in Schleicher); or that
• the disclosure was mixed with a number of positive and negative statements which could wash out the price effects of the correction.
Form and substance. Justice Scalia asked what if the 5th Circuit had simply rephrased its holding to have found that defendants rebutted the presumption of reliance by showing no movement in price. In essence, he was asking whether the difference between proving an efficient market and proving loss causation is merely semantic. Again, requiring plaintiffs to prove that prices in a market for a particular stock generally react to material information is quite a bit different (but still not easy) than requiring plaintiffs to prove that particular disclosures caused particular price changes.
The Court need not address the issue of how trial courts test for whether a market is efficient as a general matter. The Court certainly could – they could do an exhaustive look at things like the Cammer factors and review theoretical and empirical debates on finance theory. But it isn’t necessary.
Justice Alito and Speculation on Partially Inefficient Markets.David Boies did seem to flail a bit when Justice Alito asked about the beliefs of some economists that
even in a market that is generally efficient, there can be instances in which the market does not incorporate certain statements into the price of a stock.
If this were true, Justice Alito asked, shouldn’t defendants be allowed to rebut the presumption in Basic.
The Fifth Circuit in Oscar used similar logic citing a twenty year old Macey and Miller article (Oscar, 487 F.3d at 269). But, the Fifth Circuit used this argument without requiring defendants to offer any evidence at all. Under Oscar, plaintiffs essentially had to disprove that a market wasn’t inefficient with respect to certain kinds of information.
But let’s say that defendants could actually offer specific proof that a market was inefficient with respect to certain kinds of information. What would that proof look like in reality? A lack of a price movement in response to corrective disclosure has all the problems noted above: there may be other simultaneous disclosures that cloak the causal effects of the corrective statements. Or the market may also have learned about the falsehood previously. So trial courts would need lots of facts.
A fight over specific price effect of specific statements is then beginning to look a lot like loss causation. And to sort all of this out, it is beginning to look a lot like a trial on the merits at class certification. But don’t worry: Halliburton’s counsel said trial courts have powers to require more discovery.
Presumptions. Rather than turn class certification into a full blown trial, perhaps we could stick with what we have in Basic. As Justice Breyer indicated in his questions, fraud-on-the-market is about presumptions. This returns to the often overlooked passages in Justice Blackmun’s opinion in Basic on what are presumptions and why they are used. If plaintiffs can prove that market prices generally react to material information, then courts can presume that any material misstatement distorted the overall price for all investors who bought or sold during the period the misstatement was uncorrected. Without this presumption, there are not common issues to all investors.
Rather than open up a can of worms, perhaps a simple, dare-I-say Basic opinion is warranted.
In the interest of full disclosure, again, Professor Black and i joined 16 other law professors in an amicus brief arguing that the Fifth Circuit should be reversed.
TrackBack URL for this entry:
Links to weblogs that reference Halliburton Roundtable: the arguments I wish I heard: