2 AI Stocks to Buy on This Pullback

Nvidia reported a blockbuster quarter—revenue up 73% to $68 billion with record data-center sales and EPS above estimates—yet shares pulled back amid overheated Stage 1 AI expectations. The note argues the market dislocation favors "Stage 2" firms and highlights Tyler Technologies (TYL), whose municipal software franchise has ~2% annual churn, insider buys (director 1,600 shares; CAO 610 shares), a 45% Y/Y share decline and a $500 target (~49% upside). It also profiles Zscaler (ZS) as the zero-trust leader with vast real-world threat telemetry, argues Anthropic plug-ins (Claude Code) will displace some DevSecOps tools but not replace Zscaler, and assigns ZS a $260 fair value (~60% upside).

Analysis

Market structure: The durable winners are high-switching-cost Stage‑2 software and telemetry-rich cybersecurity platforms (TYL, NOW, TRI, ZS) that monetize embedded workflows and real‑time data; losers are point-solution DevSecOps/code‑scan vendors (GTLB, parts of PANW) and some Stage‑1 hardware names priced for perfection. Competitive dynamics favor incumbents with network effects—Zscaler’s threat intelligence moat and Tyler’s 2% churn create pricing power and room to cross‑sell (management cites 3→8–10 products/customer), implying steady ARPU growth over 3–5 years.

Risk assessment: Tail risks include regulation on municipal data sharing, a hyperscaler-built zero‑trust offering, or a recession-driven municipal budget squeeze that delays enterprise upgrades; probability medium, impact high. Time horizons: expect volatile re-rating in days/weeks around earnings/AI plugin launches, measurable revenue lift over 12–36 months; hidden dependencies include municipal procurement cycles and ZS’s reliance on global traffic volume as a data moat.

Trade implications: Tactical allocation should rotate capital from overbought Stage‑1 hardware into Stage‑2 software/cyber — establish concentrated longs in TYL and ZS while using small shorts in vulnerable DevSecOps names to hedge. Use option spreads to express directional views with defined risk: 9–12 month call spreads on ZS to capture the AI‑driven demand tailwind while capping premium spend.

Contrarian angles: The market is underpricing procurement inertia and service stickiness—TYL’s 45% drawdown looks overdone relative to a 6% market share and insider buys; similar to post‑2000 where infrastructure makers fell but embedded enterprise software (SAP, CRM) compounded. Unintended consequences: rapid AI tooling releases could accelerate M&A of Stage‑2 vendors (positive) or create a transient feature parity surge (negative) that compresses multiples for exposed incumbents.

Key Decisions for Investors

• Establish a 2–3% long position in Tyler Technologies (TYL) at ~ $350 per share size today; set a tactical stop‑loss at 20% and scale to 4% on a pullback to $300 or on a confirmed municipal contract streak (3+ wins/quarter).
• Open a 1.5–2% long position in Zscaler (ZS) targeting +60% in 12–24 months (model fair value ~$260); hedge by shorting PANW or GTLB at 0.6x notional to express relative DevSecOps risk. Exit or trim if ZS customer telemetry growth slows below +10% YoY for two consecutive quarters.
• Deploy a funded call spread on ZS to limit premium: buy the 9–12 month call spread sized to 0.5–1% of portfolio (example: long Jan9/Jan12 200/260 spread sizing per account risk), close on 50% of max profit or if IV rises >40% above 30‑day average.
• Reduce Stage‑1 hardware exposure (e.g., NVDA-sized positions) by 25–50% within 2–4 weeks and reallocate proceeds into Stage‑2 software/cyber; revisit re‑allocation if Stage‑1 forward earnings growth < consensus by >200bps on next two quarterly reports.