The Stocks Taking Off Now

The piece argues markets are rotating from AI applications to the infrastructure that will support large-scale AI, noting heavy capex from Microsoft, Amazon, Meta, Alphabet and Elon Musk’s effort to link 100,000 GPUs in Tennessee. It highlights InterDigital (IDCC) as an infrastructure beneficiary—sales up 28% year-over-year and a 'B' rating in Louis Navellier’s Stock Grader—and points to S&P 500 performance (2023 +24%, 2024 +23%, 2025 +16%, YTD 2026 <1%) to argue winners may shift to companies enabling AI compute and data transmission. Investors should consider repositioning toward infrastructure exposures where margins are forming.

Analysis

Market structure: The rotation favors owners of physical AI infrastructure — hyperscale cloud (MSFT, AMZN, GOOGL), networking/IP (IDCC), GPUs and data‑center builders — because rising capex converts one‑time spend into durable licensing, recurring cloud revenue and pricing power. Short‑term losers are high‑multiple, consumer‑facing AI apps where growth has been priced for perfection; expect multiples to compress by 10–30% if revenue growth decelerates. Commodities (copper, natural gas, diesel) and power markets will see incremental demand 1–3% annually where data‑center buildouts concentrate.

Risk assessment: Tail risks include hard export controls to China, a Taiwan‑stressor disrupting advanced fabs, or a hyperscaler overbuild causing a 30–50% revenue re‑pricing in interconnect/hosting within 18–36 months. Immediate (days) flows reprice names on headlines; short term (1–6 months) depends on capex guidance from MSFT/AMZN/GOOGL; long term (2–5 years) depends on ERP of power/grid upgrades and semiconductor supply cycles. Hidden dependencies: software stack efficiency (model compression) and regional power constraints can blunt hardware demand.

Trade implications: Favor infra exposure with size limits: target companies that capture recurring margins (IDCC, MSFT, AMZN, GOOGL) and use option structures to buy convexity. Pair trades that long narrow infra/IP winners vs short broad, overlevered AI app baskets will isolate margin rotation. Use 3–18 month timeframes tied to hyperscaler earnings and CHIPS/govt subsidy announcements as re‑rating catalysts.

Contrarian angles: The consensus underestimates regulatory and power bottleneck risk and overestimates smooth scaling of GPU supply; conversely the market may underprice small, high‑margin IP plays (IDCC) that can re‑rate 25–50% with a few large licensing wins. Historical parallel: Intel’s pivot shows value accrues to infrastructure owners, not the flashy apps — but mis‑timed capex cycles can flip winners into losers in 12–36 months.

Key Decisions for Investors

• Establish a 2–3% long position in IDCC (InterDigital) within 2 weeks; target +35% in 12–18 months on 28% YoY revenue growth and licensing leverage; hard stop‑loss at -18% and scale half out at +20%.
• Reallocate 3–4% from top‑decile AI app winners into cloud‑infra longs: MSFT (1.5%), AMZN (1.5%), GOOGL (1%) to capture hyperscaler capex; set trailing stop of 12% and target 12‑month upside 15–25% driven by cloud margin expansion.
• Implement a relative trade: long IDCC (1.5% notional) vs short ARKK (1.5% notional) for 6–12 months to express infrastructure vs momentum app dispersion; unwind if ARKK underperforms by >15% or if IDCC rallies >40%.
• Use options to control risk: buy a 12‑18 month IDCC call spread (buy LEAP near‑ATM, sell 30% OTM) sized to mimic a 1% equity exposure for leveraged upside; buy 3‑month 5% OTM puts on QQQ sized to protect ~30% of your tech beta (roll monthly if vol >30%).
• Monitor regulatory/capex catalysts: within the next 60–90 days, if (a) U.S./EU export controls widen to include HBM/GPU or (b) a hyperscaler announces >$10bn incremental AI capex, adjust infra longs +/-50% accordingly (reduce if controls tighten; increase if >$10bn announced).