Nvidia CEO Jensen Huang Just Delivered Fantastic News to Shareholders

Nvidia projects at least $1 trillion of sales from its Blackwell and Vera Rubin GPUs/platforms over the next 21 months through 2027, a figure that doubles its prior 2025–26 AI hardware expectations and beats Wall Street consensus. CEO Jensen Huang said Nvidia is restarting H200 processor sales to China after several quarters of restrictions, though the U.S. may take ~25% of related revenue and Nvidia previously took a $5.5 billion charge tied to H200 inventory and commitments. The resumption could prompt analysts to raise revenue and EPS estimates and supports the stock given NVDA trades below ~22x forward earnings.

Analysis

Market reaction is focusing on headline growth — the second-order winners are the foundry and advanced-equipment suppliers that must scale capacity and test/pack partners that will see elongated lead times and order phasing over the next 12–24 months. That creates a choke-point dynamic: incremental demand can drive outsized pricing power for constrained suppliers even as OEMs and hyperscalers face lumpy delivery and inventory timing risk.

Geopolitical reopening to a large, previously restricted market creates an atypical mix of upside to top-line growth and downside to margins because of any government-mandated revenue sharing, compliance costs, and increased channel/distributor complexity. Expect initial revenue recognition to be front-loaded (channel fill, prepayments) followed by a 3–9 month normalization or destocking, creating a volatile EPS profile through the next two fiscal quarters.

Competitors and ecosystem players face bifurcated impacts: incumbents focused on GPUs capture outsized share of near-term AI capex, while CPU-specialists and alternative AI-silicon vendors will accelerate product roadmaps and customer price concessions to defend wallet share. Over a 12–36 month horizon, domestic suppliers in constrained markets will receive preferential procurement, raising substitution risk for western vendors unless they lock into long-term supply agreements.

Primary near-term catalyst risk stems from policy reversal or enforcement tightening; secondary catalysts include foundry capacity ramp rates and evidence of durable adoption (multi-year purchase commitments vs one-time orders). The consensus likely underweights margin dilution and front-loaded channel effects while overestimating how quickly incremental sales will convert to sustainable operating profit.

Key Decisions for Investors

• Buy NVDA 12–24 month LEAP calls ~10–25% OTM (size = 3–5% NAV). Rationale: asymmetric upside from share gains and multiple expansion if demand sustains. Plan to take profits at 2.5x premium or roll if conviction unchanged; cap loss at 50% premium decline.
• Establish a downside hedge: purchase a small NVDA put spread (same expiry as LEAPs) sized to cover ~30–40% of the long position cost. Use this to protect against a rapid regulatory reversal or Chinese destocking event within the next 6–9 months.
• Pair trade: long NVDA equity / short INTC equity on a dollar-neutral basis (initial size 2–4% NAV each leg). Rationale: capture secular CPU-to-GPU share shift while hedging market beta. Trim the short if NVDA/INTC performance divergence exceeds 30% in 3 months or if Intel announces a credible hyperscaler design win.
• Event trigger rule: add to longs on confirmed multi-year purchase agreements or sustained order flow evidence over two consecutive quarters; reduce exposure if channel inventory indicators (third-party reseller sell-through, semiconductor equipment lead indicators) show >20% build vs expected demand within 90 days.