Hedge fund founder Eric Jackson announced Tuesday that he has taken short positions in five software companies - Asana, Salesforce, Five9, DocuSign, and Atlassian - arguing his research shows firms that most frequently invoke artificial intelligence in earnings calls tend to underperform their peers.
Jackson - who previously ran an activist campaign with Opendoor Technologies last year - released findings from a study that initially examined 716 earnings calls across software companies. The work indicated that companies in the top quintile for AI-related language density underperformed those in the bottom quintile by 5.4 percentage points over the subsequent 90 days. The result carried a p-value of 0.043.
He characterized the observed pattern as the "AI Paradox," saying management teams often lean on AI narratives as a response to competitive threats posed by AI automation. In addition to the broader sample, Jackson's analysis specifically covered 147 earnings call transcripts from 21 enterprise software companies.
Jackson highlighted Chegg as an illustrative case. According to his account, the education-technology company declined from a roughly $14 billion valuation in February 2021 to trading at $0.67 per share after the launch of ChatGPT affected its homework-help subscription business. Following Chegg's Q1 2023 earnings call, the stock fell 48% in a single day, a drop Jackson pointed to as consistent with the pattern his research describes.
The research asserts enterprise software vendors may be pressured by AI agents capable of handling structured, repetitive workflows without traditional per-seat licensing models. Jackson referenced Klarna's choice to replace Salesforce with an internally developed AI system and to shut down 1,200 software vendors as an example of that shift.
Market reaction to Jackson's short positions was muted in directional terms. On Tuesday, Atlassian (NASDAQ:TEAM) rose 1.7%, Five9 (NASDAQ:FIVN) gained 0.5%, Salesforce added 0.4%, DocuSign (NASDAQ:DOCU) increased 0.2%, and Asana (NYSE:ASAN) climbed 2.9%.
Summary
Eric Jackson has publicly disclosed short positions in five software companies after publishing a study linking high AI-related language on earnings calls to subsequent underperformance. His work, covering a broad set of earnings calls and a focused sample of enterprise software transcripts, identifies a measurable near-term performance gap and presents examples intended to illustrate how AI-driven disruption can affect company valuations and revenue models.
Key points
- Jackson shorted Asana, Salesforce, Five9, DocuSign, and Atlassian based on his research findings.
- Across 716 software-company earnings calls, firms in the top quintile of AI language density underperformed the bottom quintile by 5.4 percentage points over 90 days (p = 0.043).
- The analysis included a focused review of 147 earnings call transcripts from 21 enterprise software companies, and cites Chegg and Klarna as examples of the dynamics described.
Risks and uncertainties
- Market reaction can differ from the research signal - all five targeted stocks traded higher the day Jackson disclosed his positions, showing short-term price moves may not align with the study's findings.
- The research describes an observed association rather than a guaranteed outcome; underperformance measured over 90 days is a statistical result subject to variability and the stated p-value.
- Examples cited, such as Chegg and Klarna's actions, illustrate potential pathways but do not establish that all companies discussing AI will experience similar valuation outcomes.
Jackson's report frames AI-related corporate commentary as a potential warning sign rather than proof of impending decline. Investors and market participants will likely weigh his findings against other fundamental and market signals when assessing enterprise-software valuations and competitive risks.