Eric Jackson shorts five software stocks citing ’AI Paradox’ By Investing.com

Hedge fund founder Eric Jackson announced short positions in Asana, Salesforce, Five9, DocuSign and Atlassian after analysis showing companies in the top quintile of AI language density underperformed the bottom quintile by 5.4 percentage points over the subsequent 90 days (p=0.043). His research covered 716 earnings calls (147 transcripts across 21 enterprise software firms) and cites Chegg’s collapse and Klarna replacing Salesforce as precedents. Despite the short calls, the five named stocks traded modestly higher on Tuesday (Asana +2.9%, Atlassian +1.7%, Five9 +0.5%, Salesforce +0.4%, DocuSign +0.2%).

Analysis

AI narrative intensity on calls is acting as a real-time signal of structural vulnerability for vendors whose economics depend on per-seat licensing and predictable human workflows. Firms with high seat density and low per-customer technical differentiation face two compressive forces over 12–36 months: (1) unit economics arbitrage as task-based AI agents undercut per-seat pricing, and (2) procurement-level consolidation as buyers internalize models and push vendors to feature-bundled, lower-margin contracts. Watchable leading indicators are falling seats per account, widening gap between billings and recognized ARR, and step-downs in net revenue retention; a sustained 3–6 month drift in any of these metrics typically precedes a >15% re-rating in similar software names.

Secondary beneficiaries are cloud infra and data-platform providers who capture the marginal spend when customers build internal agents: compute, embeddings stores, and MLOps tooling see stickier, higher-margin flows than shrinking seat-license payments. Conversely, mid-tier ISVs, point SaaS vendors, and systems integrators that monetize repetitive workflows are at highest risk of revenue displacement and rapid vendor consolidation — expect increased M&A of underperforming assets at distressed multiples within 6–18 months. Monitoring procurement RFP language for “in-house model” or “API-first” is an early procurement signal.

Risk/catalyst framework: near-term price action can be dominated by sentiment and short-covering (days–weeks), while structural ARR erosion plays out over quarters. Reversal scenarios include vendors proving task-level AI materially increases switching costs (moving to outcome-based pricing), regulatory limits on model deployment that raise costs for internal builds, or a large incumbent bundling AI as a retention moat. Set 3-, 12-, and 36-month monitoring windows: tactical means (earnings/commentary) for the former, and customer-level KPIs for the latter.

Key Decisions for Investors

• Initiate a 3–6 month tactical short basket (equal-weighted): ASAN, DOCU, FIVN, TEAM. Target 20–30% downside vs entry with a hard stop at +15% adverse move; size to limit portfolio-VaR impact. Rationale: high seat-based exposure + visible procurement shift risk; hedge with 1/3 notional long on cloud infra (e.g., AMZN or MSFT) to capture migration of spend.
• Pair trade (6–18 months): Short CHGG equity or buy 12-month puts (limit order) vs Long KLAR via 12–18 month call spread (buy KLAR 1.5x ATM call / sell 2.5x call). Aim for ~2:1 skewed payoff where CHGG downside captures direct AI substitution risk and KLAR captures disintermediation gains or payments-led monetization upside.
• Event-driven fragmentation trade: Buy out-of-the-money 3–6 month puts on FIVN (contact-center automation exposure) funded by selling weeklies on highly liquid, less-exposed software names; profit if procurement shifts accelerate. Close on first evidence of sustained >5% QoQ seat reduction or a major customer publicizing internal build.
• Monitor fundamentals as triggers: reduce shorts or flip to long if a target reports NRR >110% with rising ARPA or announces proprietary AI that demonstrably raises switching costs (e.g., embedded vertical logic, exclusive data partnerships). Maintain explicit stop-loss and size limits given short-squeeze risk in small/mid-cap software names.