William Blair reiterates Outperform on Nvidia stock after GTC conference

Nvidia is trading with a $4.42 trillion market cap after 65% revenue growth over the last 12 months and InvestingPro flags the stock as below its Fair Value while 33 analysts have raised earnings estimates. William Blair reiterated an Outperform after the GTC conference, citing system-level co-design and rack-level performance advantages; Rosenblatt raised its price target to $325 (Buy) while Wolfe ($275), Bernstein ($300) and Stifel ($250) maintained positive views. Analysts highlighted new silicon (e.g., Groq LPX) and strong order visibility for Blackwell and Rubin, with Rosenblatt projecting up to $1 trillion in revenue from those products between 2025–2027, reinforcing upside potential for NVDA.

Analysis

Nvidia’s system-driven edge creates defensive dynamics for incumbents and suppliers: customers buying at rack/system level lock in long bill-of-materials windows and create lumpy, multi-quarter procurement cycles. That amplifies upstream demand volatility for specialized components (HBM, high-end PSUs, interconnect ASICs), which can produce asymmetric margin tailwinds for component suppliers while compressing spot market availability for general-purpose foundry capacity.

The commercialisation of high-throughput inference shifts value from raw silicon to validated system integration and software stack capture, raising switching costs for hyperscalers but lowering marginal value for smaller GPU-focused rivals. This bifurcation favors vertically integrated OEMs and hyperscalers that can absorb higher ASPs while creating a narrow moat for firms unable to fund equivalent systems engineering investments.

Timing and risk separation is critical: expect near-term (0–3 months) headlines and order beats to squeeze implied vol and re-rate equities, while the 6–24 month horizon is where competition, regulation and supply constraints determine realized margins. Key tail risks that would reverse the rationale are rapid foundry capacity expansion for alternative architectures, a material China-access restriction, or an abrupt hyperscaler capex pause driven by macro weakness.

Consensus currently prizes top-line growth; what’s underpriced is the probability of margin normalization as customers extract price concessions once product sets mature. That argues for structures that capture convex upside around execution beats while limiting exposure to a multi-year margin mean reversion scenario.

Key Decisions for Investors

• Buy a 12-month NVDA call spread (buy ~0.35-delta call, sell ~0.12-delta call) sized to 1–2% NAV to capture platform upside while capping premium; target 30–60%+ equity upside, max loss = premium paid, breakeven ~25–30% underlying move.
• Initiate a 6–9 month long SMCI position (equity or 0.40-delta calls) sized 0.5–1.5% NAV to play OEM/order flow capture from systemization; set a tactical stop at -25% and take-profit tranche at +60%, hedge by selling short-dated NVDA calls (to fund premium) if running net long NVDA exposure.
• Buy 3-month NVDA protective puts (~0.15–0.20 delta) equal to 1–1.5% NAV as insurance against headline-driven 15–30% drawdowns around near-term releases/earnings; cost is limited premium and preserves upside participation elsewhere in portfolio.
• Short APP via a 9–12 month put spread (sell 0.25-delta put, buy 0.10-delta put) or small outright short (<=0.5% NAV) to express skepticism on durable monetization of AI halo into mobile ad CPMs; target 30–40% downside, risk limited to spread premium or defined to NAV sizing.