AI agents could easily send college grad unemployment over 30%, ServiceNow CEO says

Bill McDermott warned AI-driven productivity gains could push unemployment for new college graduates into the mid-30s% within a few years; the NY Fed estimates recent-graduate unemployment ~5.7% at end-2025 and underemployment at 42.5% (highest since 2020). Large tech firms are already cutting jobs—Block plans to cut nearly half its workforce, Atlassian will cut ~10% after its stock fell ~54% YTD—and ServiceNow says it removed ~90% of prior human customer-service use cases, enabling lower hiring and higher free cash flow. CEOs at Palantir and Amazon have likewise signaled aggressive revenue/headcount plans tied to AI, implying sector-level headcount reductions and cost takeout ahead.

Analysis

Winners will be platform and consumption-priced vendors that convert headcount-driven workflows into measurable FCF expansion; vendors that charge per-seat or per-user risk revenue compression as enterprises automate away low-value seats. ServiceNow is positioned to capture high-retention workflow automation spend and expand gross margins, while collaboration and developer-tool vendors with user-count economics (e.g., Team-style names) face dual pressure on growth and pricing power.

Second-order effects: reduced entry-level hiring depresses near-term consumption in services tied to early-career wages (recruiting firms, office space sublets, corporate training vendors), tightening demand in cyclical consumer categories over 6–18 months. At the same time, large-scale automation raises demand for AI infrastructure and observability (cloud compute, model monitoring), creating a bifurcated market where infrastructure names can outgrow application-layer names even as some apps consolidate.

Key risks and catalysts — timeline matters. Near-term (days–months) catalysts are layoff announcements, earnings beats/misses tied to automation ROI, and employment prints; medium-term (6–24 months) catalysts are enterprise adoption curves, pricing-model pivots (per-seat → consumption), and regulatory pushback on mass automation. Tail risks include regulatory constraints on workforce automation, an AI productivity plateau, or a surge in labour activism that forces rehiring; any of these could materially reverse positioning.

Key Decisions for Investors

• Long NOW via a 9–12 month call spread (buy 12-month ITM call / sell 20–30% OTM call) sized 3–5% net portfolio exposure. Rationale: capture FCF/GM expansion as platform adoption accelerates; target asymmetric upside of ~+30–40% with defined downside equal to premium paid.
• Short TEAM tactical via 3–6 month put spread (sell 6–8% OTM put / buy deeper 15–20% OTM put) size 2–3% notional or outright 2% position short-stock for momentum. Rationale: user-count monetization risk and margin compression; reward if earnings guidance weakens or layoffs persist. Maximum loss limited to spread width or stock movement if shorted.
• Pair trade: long NOW / short TEAM equal notional for 3–12 months to isolate automation-adoption vs per-seat monetization risk. Rationale: hedges macro beta and expresses structural rotation into consumption-priced automation platforms; rebalance after earnings or major product announcements.
• Small conviction long PLTR LEAP (9–18 month calls) at 1–2% notional as a levered hedge to AI infrastructure winners. Rationale: optionality on enterprise model-deployment demand — asymmetric upside if customers accelerate large-scale deployments amid corporate headcount cuts.
• Protective hedge: buy a small-cap weighted basket of 3–6 month put spreads on AMZN/large conglomerates (total 1–2% notional) to guard against demand shock if entry-level unemployment materially pulls consumer spend. Rationale: tail insurance with limited cost if automation-driven consumption declines faster than expected.