A young analyst is most likely typing the final paragraph of a report that could reduce a company’s market value by hundreds of millions by Tuesday morning in a small Vancouver office with exposed brick walls and an excessive number of monitors. This is a close-up view of the short-seller comeback. Not glitzy. not always broadcast on television. However, it is becoming more and more significant, particularly as the AI boom floods public markets with businesses whose claims exceed their earnings by a sizable and dubious margin.
This movement began in late 2010 when Sahm Adrangi of Kerrisdale Capital posted a brief thesis about China Education Alliance, a Chinese company with a New York listing and impressive statistics. A YouTube video depicting CEU’s flagship campus in Harbin as an eerily empty, unfurnished structure was another aspect of the research that stood out. The stock fell. Adrangi’s fund made money. And a template was created without anyone really planning it.
It is now more difficult to contain what Adrangi and a few other like-minded analysts began back then. By publishing thorough, sometimes brutal takedowns of public companies and forcing them out through social media before the target even knows what’s coming, anonymous research firms like Aurelius, Viceroy, and Iceberg have made careers—and occasionally enemies. After witnessing the growth of these campaigns after 2019, the Canadian Securities Administrators initiated a formal consultation process, posing the difficult question that regulators continue to ponder: at what point does exposure to legitimate fraud turn into market manipulation?
The current AI small cap market may be the most target-rich time period that activist short sellers have ever encountered. Numerous thinly traded companies have rebranded themselves around artificial intelligence on North American exchanges, sometimes with little more than a press release and a new tagline on their homepage. Some have genuine goods. Many appear to be riding the wave long enough to draw in retail investors who lack the resources or time to decipher financial filings. As it happens, short sellers do have that time.
Over the past two or three years, there seems to have been a change in the dynamic. In contrast to the initial wave of short-seller campaigns, which targeted Chinese reverse-IPO frauds—businesses that had successfully gained access to American exchanges through backdoor mergers—the current campaigns are more domestic in nature and have a greater emphasis on the artificial intelligence industry. The strategy is similar: examine the financial data, if at all possible, visit the physical locations, cross-reference the purported customer relationships, and then publish all of this information in a report that is meant to be read, shared, and trusted. Now, reach makes a difference. By noon on a Tuesday, a brief thesis could become popular.
Regulators are still genuinely unsure of how difficult it will be to resist. The conflict was openly acknowledged in the CSA’s consultation paper. Activist short sellers contend that they are the first line of defense against fraud, spotting what auditors overlook and what regulators take too long to investigate. It makes sense that businesses have a different perspective. Before management has a chance to react, a single well-timed report can destroy a share price, eliminating shareholders who had nothing to do with the alleged wrongdoing. The speed at which brief reports are processed is rarely matched by the legal system’s ability to determine who was correct and how quickly.
It’s difficult to ignore a recurring pattern when observing this develop in the AI small cap space in particular. A business announces a collaboration with a significant technology platform. Shares rise. Three weeks later, a brief report questions the validity of the partnership, the revenue terms, and the veracity of the CEO’s purported background. The business reacts angrily. The stock swings. The truth eventually becomes clear after a few weeks or months. The short sellers are not always correct. The issuer community is not, more often than they would like to acknowledge.
The more fundamental question is whether this type of adversarial research is a strength or a weakness, a question that neither markets nor regulators have satisfactorily addressed. The short sellers will claim that markets depend on them and that without them, dishonest businesses would be able to raise more money from more people for longer. The businesses will tell you that the threat is disproportionate and that years of good work can be destroyed by a single anonymous report with questionable sourcing before anyone looks into it. Frustratingly, both sides have a point. Furthermore, the AI boom has made the debate more difficult to settle than before due to its hype cycles, exaggerated valuations, and real challenges in determining what these technologies can truly accomplish.
