Nvidia Stock Looks ‘Cheap,’ Analysts Say

Truist Securities analyst William Stein reiterated a buy rating on Nvidia and raised his price target to $275 from $255, citing that AI infrastructure semiconductor stocks “remain cheap.” Nvidia shares closed up 3.9% at $180.99 on the news, underscoring continued investor appetite for AI chip exposure and implying further upside if analysts and demand trends persist.

Analysis

Market structure: Nvidia is the clear incumbent beneficiary — accelerating data‑center GPU demand and China approvals increase TAM visibility and pricing power; expect NVDA revenue mix to shift +5–10pp toward hyperscaler direct sales over 12–18 months, pressuring smaller GPU vendors and traditional CPU suppliers. Supply remains constrained by TSMC capacity and packaging lead times, implying near‑term selling discipline and sustained ASPs rather than a rapid price war. Cross‑asset: a sustained AI rally will push equity risk premia lower (risk‑on), modestly steepen US yields (10–25bp) and lift equity implied vols in semis; USD may soften slightly on risk‑on, supporting EM assets and copper demand for datacenter buildouts.

Risk assessment: Tail risks include US/China export restrictions reversal or geopolitical disruption in Taiwan (high impact, <10% prob.), rapid emergence of in‑house hyperscaler silicon (Amazon/Google) eroding NVDA share (20–30% medium‑term share risk), or a demand “show‑me” pause if AI monetization lags. Immediate (days): momentum and options gamma dominate; short term (weeks–months): order cadence, supply shipments and Q/Q guide; long term (years): architecture wins and ecosystem lock‑in. Hidden deps: NVDA’s growth hinges on TSMC capacity, CCL yield curves, and large cloud deals (OpenAI/Amazon) — watch vendor contract disclosures and foundry utilization.

Trade implications: Tactical overweight NVDA via concentrated but size‑capped exposure — use LEAPs or call spreads to control downside; consider relative shorts in names with weaker AI moats (Oracle, Broadcom, CoreWeave). Options: prefer vertical call spreads (buy Jan‑2027 240/360 call spread) to cap premium, and sell near‑dated covered calls after 10% rallies to harvest IV. Rotate 3–6 month exposure from legacy enterprise software and non‑AI hardware into semis and cloud infra (AMZN, GOOGL) to capture capex cycle.

Contrarian angles: Consensus underestimates hyperscaler vertical integration risk and overestimates perpetual multiple expansion; NVDA’s run to $275 PT is plausible but pricing already embeds aggressive TAM growth (market assumes >30% CAGR next 3 years). Reaction is partly underdone on supply constraints but possibly overdone on long‑dated profitability if competitors or regulation bite; historical analogue: high‑margin hardware cycles (GPU in 2016–18) saw sharp mean reversion when software monetization lagged. Unintended consequence: concentrated AI spending could trigger antitrust/competition responses or capex pullbacks if ROI benchmarks aren’t met, creating a sharp re‑rating event.

Key Decisions for Investors

• Establish a 2–3% portfolio long in NVDA within 2 weeks via a mix of shares (50%) and Jan‑2027 240/360 bull call spreads (50% notional) — scale into dips to 165–175; trim half position at 260–280 and fully exit on a 35% drawdown from entry.
• Implement a pair trade: long NVDA 2% vs short AVGO 1.5% (net long beta ≈0.5) to express AI infra outperformance vs legacy semiconductor consolidation; rebalance if AVGO underperforms by >12% or NVDA exceeds target 275.
• Use options to express asymmetric upside: buy Jan‑2027 NVDA 240/360 call spreads sized to risk 0.5–1% of portfolio; after any >10% one‑day NVDA move, sell 30–45 day 25–35 delta calls (covered or naked depending on size) to monetize elevated IV.
• Trim 2–4% exposure to Oracle (ORCL) and CoreWeave (CRWV) over the next 60 days and redeploy into AMZN/GOOGL (1–2% each) to capture cloud stack capture; stop redeploy if ORCL reports ≥3% beat in cloud ARR or if NVDA guidance cuts revenue by >5% vs consensus.