Raymond James raises Nvidia stock price target on GPU sales outlook

Raymond James raised its NVIDIA price target to $323 from $291 and kept a Strong Buy, citing management's outlook for $1 trillion of cumulative GPU sales through 2027; NVDA trades around $180 and is flagged as undervalued. Multiple brokers also issued bullish actions (Truist $287 PT, Bernstein $300 PT; William Blair and Stifel reiterated positive ratings), 33 analysts have raised earnings estimates, and NVIDIA announced a Qnity collaboration — all supporting upside for NVDA and supply‑chain names (Coherent, Lumentum, NetApp, Everpure, Dell, Supermicro, HPE).

Analysis

The market is treating GPU-led AI inference as a multi-year, structurally compounding demand stream that pulls revenue through the optical, storage and system OEM supply chains. Expect biggest near-term margin expansion at vendors that control system-level integration and optical interconnect IP, while discrete component makers (HBM, substrates, advanced packaging) will see boom-bust dynamics driven by foundry and substrate lead times.

A key second-order effect: higher-density inference racks materially change server economics — more optical ports per chassis, higher HBM counts and denser power delivery — which concentrates incremental TAM into a smaller group of system and optical suppliers and amplifies their order-volatility vs. legacy server suppliers. This creates a two-tier market where a handful of winners can re-rate quickly while broad hardware indices lag even as aggregate capex grows.

Tail risks that would reverse the trade include demand elasticity from end customers (AI spend trimming if ROI diverges from expectations), accelerated competition from custom accelerators or a renewed export-control shock; both could cause order float and inventory digestion over 2-6 quarters. Near-term catalysts to watch: quarterly guidance cadence (days), product ramp cadence and supply commentary (months), and multi-year order visibility or cancellations (quarters-to-years).

Key Decisions for Investors

• Long NVDA (directional, 12–24 months): buy Jan-2027 220/420 call spread sized to 1–2% risk budget. Rationale: captures multi-year inference adoption while capping downside to premium; target 2.5–3x return if NVDA sustains unit and ASP momentum, max loss = premium.
• Tactical optics exposure (6–12 months): buy calls on COHR and LITE or 6–12 month outperformance swaps (equal $ notional). Rationale: optical port growth per rack is a high-leverage proxy to GPU server deployments; allocate small, active positions and trim on 30–50% rallies.
• Storage/system pair (6–12 months): long NTAP, HPE (scale/higher margin) vs. short broader legacy server exposure (e.g., large diversified OEM) — or simply overweight NTAP vs. DELL through stock or call purchases. Rationale: storage offload and NVMe/SSD density reprice faster than low-margin OEM cycles; target 20–40% relative upside, stop-loss 12–15%.
• Tail hedge (0–6 months): buy 3–6 month ATM puts on NVDA or a SOXX put spread sized as 25–50% notional hedge against a 20–30% drawdown. Rationale: protects concentrated AI exposure against inventory normalization, macro shock or regulatory export moves; accept cost as insurance (1–3% portfolio drag).
• High-risk optionality (9–18 months): small allocation to Q (collaboration exposure) via long-dated calls to capture upside if partnerships drive materials/stack wins. Size <0.5% risk budget; expect binary outcomes and high volatility.