The Zacks Analyst Blog NVIDIA, AT&T and Amgen

Zacks highlights research on NVIDIA, AT&T and Amgen showing YTD outperformance for all three: NVIDIA +35.9% vs. semiconductor industry +34.2% driven by AI demand for Hopper/Blackwell GPUs and Ada RTX workstation ramp but constrained by Blackwell supply, rising AI system costs and U.S.-China tech tensions; AT&T +17.6% vs. wireless industry +3.7% supported by postpaid momentum, fiber expansion (targeting >50M locations by 2030) and a 70% Open RAN traffic goal by late 2026, offset by wireline losses and high debt; Amgen +32.5% vs. biomedical industry +19.9%, with Q3 beats and growth from products like Evenity, Repatha, Blincyto and new launches (Tavneos, Tezspire) while facing pricing pressure, biosimilar competition and LOE risks. Investors should weigh strong top-line and product catalysts against supply constraints, margin pressure and competitive/regulatory headwinds.

Analysis

Market structure: NVIDIA (NVDA) is the primary beneficiary of escalating generative-AI demand; constrained Blackwell GPU supply implies pricing power and upside to data-center ASPs but caps near-term revenue growth if wafer/packaging cannot ramp. AMD (AMD) is the obvious competitive loser near-term (market-share pressure in NVDA’s high-end GPU segment), while AT&T (T) and fiber suppliers benefit from predictable long-cycle capex and steady ARPU upgrades; legacy wireline peers face margin squeeze. Cross-asset: stronger tech sentiment should tighten corporate credit spreads (compress HY by 10–30bps on a sustained AI rally), push equity vols lower (NVDA IV likely remains elevated near earnings), and increase electricity/copper demand modestly for data-center expansion.

Risk assessment: Tail risks include abrupt US-China export controls (50%+ downside shock to NVDA GPU revenues vs baseline over 6–12 months), semiconductor-capacity shocks at TSMC, and regulatory limits on AI product monetization. Short-term (days–weeks) risk centers on inventory and Blackwell supply updates; medium (3–12 months) risks are margin erosion from rising AI system costs; long-term (2+ years) risks include LOE and biosimilar pressure on AMGN. Hidden dependency: NVDA revenue growth is tightly coupled to TSMC capacity allocation and enterprise AI software adoption curves.

Trade implications: Establish a tactical 1–1.5% long NVDA via 3–6 month call spreads (buy 1x ATM, sell 1x 20–30% OTM) to cap cost while capturing continued AI upside; pair that with a 1% short AMD cash-secured put or short equity exposure to express relative share shifts. Add a 2% dividend-income position in T (buy shares and sell 1–2 month covered calls at +6–8% OTM) to harvest yield while capping drawdown; for AMGN, consider selling 1–2% 12-month puts 10% below spot to collect premium given pipeline uncertainty but attractive downside yield. Reduce long-duration government bond exposure by 25–50bps duration if reallocating into cyclicals.

Contrarian angles: Consensus may underprice AT&T’s fiber optionality—if US subsidy/access programs accelerate, T could re-rate by 10–20% over 12–24 months; conversely NVDA’s valuation is pricing near-term perfection—if Blackwell supply eases and competition from AMD accelerates, prepare to trim after any 30–40% rally. Historical parallel: tech hardware booms often see supply-driven constraints that limit revenue even as ASPs rise (2000s GPU/ASIC cycles); unintended consequence is faster capex inflation for customers, which would slow enterprise purchase cadence and create a demand trough 6–12 months out.

Key Decisions for Investors

• Establish a 1–1.5% position in NVDA via a 3–6 month bull call spread (buy ATM, sell 20–30% OTM) to capture AI/data-center upside while limiting premium outlay; set a take-profit at +30–40% and stop-loss at -15% from entry.
• Implement a relative-value pair: short AMD equal notional to 60–80% of NVDA long size (cash equity or 2–3 month puts) to express high-end GPU share migration risk; reassess after any AMD product roadmap announcement within 30 days.
• Buy 2% of portfolio in AT&T (T) for yield and fiber optionality and fund by selling 1–2 month covered calls at +6–8% OTM; if T trades 15% above entry within 12 months, trim to 1% holding.
• For AMGN, sell 12-month cash-secured puts at ~10% below current spot to collect premium and establish position only if assigned; risk/reward favors ~8–12% annualized return if no major LOE surprises within 12 months.
• Reduce portfolio duration by ~25–50bps and shift 3–5% into select semiconductor equipment and fiber infrastructure suppliers (names with >50% revenue exposure to data-center/fiber capex) to capture sector rotation while hedging rate-sensitivity.