Short and Distort

Anonymous political speech has a celebrated history (Publius, 1787) and has long enjoyed strong protections under the U.S. Constitution.[1] But there is a dark side to pseudonymity: Fictitious identities can wreak havoc in financial markets. A large literature in economics examines why markets are vulnerable to rumors and information-based manipulation (Benabou & Laroque, 1992; Van Bommel, 2003; Vila, 1989). In a review of this body of work, Putnins (2012) emphasizes the importance of reputation: “if market participants are able to deduce that false information originated from a manipulator, the manipulator will quickly be discredited and the manipulation strategy will cease to be profitable.” Pseudonymity undermines the disciplinary effect of reputation, allowing manipulators to exploit investors’ trust, profitably distort stock prices, and switch fictitious personas with impunity.

In my article, available here, I show how pseudonymity undermines reputational accountability in financial markets. I examine 2,900 articles criticizing mid- and large-cap firms and published on a website, Seeking Alpha, and show that pseudonymous ones are followed by stock-price declines and sharp reversals, leading to over $20.1 billion in mispricing. I employ propensity-score matching between pseudonymous and real-name attacks and use a triple-difference design to show abnormal put options trading with publication. On the day of publication, the open interest and volume of put options written on the target of a pseudonymous article are higher than call options. While I cannot prove that the pseudonymous author is trading, the universe of potential traders is small: Only that author, his or her tippees, or possibly the Seeking Alpha editorial staff,[2] know an article is forthcoming.

During the second to fifth day following publication, the open interest and volume of call options written on the target of a pseudonymous article are higher than put options. Both calls and puts follow parallel trends in the preceding days, strengthening the causal interpretation of the divergence in open interest and volume. Following Cremers & Weinbaum (2010), I also show these periods are characterized by deviations from put-call parity that are indicative of informed trading. A textual analysis suggests that provocative article content is unlikely to be driving these price reversals. The words and phrases correlated with pseudonymous authorship do not refer to fraud or similar evocative improprieties.

Fox et al. (2018) show that “liquidity suppliers will increase their spreads to compensate for the prospect of losing money to misstatement manipulators.” I test this proposition by examining how market makers adjust bid/ask spreads in anticipation of informed buying during the reversal period. While the publication of the article comes as a surprise to market makers, they can anticipate the possibility of selling to an informed buyer who purchases in anticipation of a post-publication price correction. I show that a strong negative abnormal return on the publication day is linked to an increase in bid/ask spreads until two days post-publication, when call options trading is expected to commence.

More importantly, I directly test the predictions of the theoretical literature on reputation (Benabou & Laroque, 1992; Van Bommel, 2003; Vila, 1989). I show that pseudonymous authors manipulate markets when they are perceived as trustworthy, i.e., when the author had few reversals in the past or has no history. Pseudonymous authors disappear after the market realizes fraud is taking place, enabling them to switch to a new identity. Finally, switching pseudonymous identities leaves subtle traces of writing style detectable using stylometry, a method of authorship attribution developed in computational linguistics.

Pseudonymous attacks pose unique challenges for securities law. A factual misstatement made with scienter gives rise to liability under Section 10(b) of the Securities Exchange Act of 1934. But pseudonymous attacks often consist of murky opinions rather than express factual claims, and it is difficult to establish intent to deceive without knowing the identity of the pseudonymous author. Moreover, many courts have resisted the notion that open-market transactions can constitute market manipulation in violation of Section 9 of the Securities Act of 1933, even with proof of intent to distort prices.

The SEC very recently has taken enforcement action against a short-and-distort scheme under a misstatement theory in the matter of Ligand Pharmaceuticals.[3] The SEC has alleged that hedge-fund adviser Gregory Lemelson and Lemelson Capital Management sought to manipulate the price of Ligand’s stock by “orchestrating a public campaign” against Ligand consisting of false statements of material facts “intended to shake investor confidence in the company, drive down the price of Ligand’s stock, and consequently, increase the value of Ligand’s short positions.”[4] The Ligand case does not involve allegations of derivatives trading, which may strengthen the inference of manipulative intent, indicating both that the SEC may be taking a more aggressive stance against short-and-distort campaigns as well as suggesting that cases with manipulative options trading may be especially ripe for SEC review.

The key to this story is online intermediaries like Seeking Alpha, who secure fictitious accounts and keep them from being hijacked by anonymous impersonators. As gatekeepers of the link between pseudonymous accounts and underlying authors, these intermediaries are well- suited to punish systematic manipulators without chilling pseudonymity itself. Rather than prohibiting identity-switching outright, regulators can hold online intermediaries accountable for tolerating market manipulation via pseudonymous attacks.

ENDNOTES

[1] In the words of Justice Stevens, “Anonymity is a shield from the tyranny of the majority.” McIntyre v. Ohio Elections Com’n, 514 U.S. 334, 357 (1995) (citing J. S. Mill, On Liberty, in On Liberty and Considerations on Representative Government 1, 3-4 (R. McCallum ed. 1947)).

[2] Seeking Alpha strictly prohibits editors from trading ahead of a forthcoming article, see https://seekingalpha.com/page/seeking-alpha-conduct-and-investment-policy, and there is no evidence that they are doing so.

[3] SEC v. Lemelson, Case No. 1:18-cv-11926, D. Mass, https://www.sec.gov/litigation/complaints/2018/comp- pr2018-190.pdf.

[4] Id.

This post comes to us from Professor Joshua Mitts at Columbia Law School. It is based on his recent article, “Short and Distort,” available here. The author consults on regulatory and litigation matters related to short-and-distort campaigns, but has no financial interest in any of the cases studied in the article.

Leave a Reply

Your email address will not be published. Required fields are marked *