Anonymous stock-market manipulators behind $20B of "mispricing" can be tracked by their writing styles

In a new Columbia Law and Economics Working Paper, Columbia Law prof Joshua Mitts uses "stylometry" (previously) to track how market manipulators who publish false information about companies in order to profit from options are able to flush their old identities when they become notorious for misinformation and reboot them under new handles.

Stylometry is a field of text analysis that seeks to identify authors by stylistic quirks, including word-choices, punctuation habits, and subtler cues like sentence-length and structure.

Mitts studied 2,000 "attacks" published on the finance site Seeking Alpha, showing that the scammers involved were shedding old identities when their financial analysis proved to be incorrect -- and also showing that someone (maybe the scammers, maybe the Seeking Alpha editors, maybe someone else) -- was making a killing by buying options before the publication of erroneous data.

Mitts uses stylometry to identify when multiple consecutive pseudonyms seem to belong to the same anonymous author, but he doesn't actually attempt to unmask the author, though the same stylometry techniques could produce evidence, if not proof, of the scammers' identities.

Pseudonymous attacks on public companies are followed by stock price declines and sharp reversals. I find these patterns are likely driven by manipulative stock options trading by pseudonymous authors. Among 1,720 pseudonymous attacks on mid- and large-cap firms from 2010-2017, I identify over $20.1 billion of mispricing. Reputation theory suggests these reversals persist because pseudonymity allows manipulators to switch identities without accountability. Using stylometric analysis, I show that pseudonymous authors exploit the perception that they are trustworthy, only to switch identities after losing credibility with the market.