Nvidia's $1T projection, Trump's Cuba threats, 1-hour Amazon deliveries and more in Morning Squawk

Nvidia projects $1 trillion in orders for its Blackwell and Vera Rubin systems through next year (about double last year’s projection) and announced multiple product and platform launches, contributing to a ~1.7% share gain after Jensen Huang's GTC keynote. Oil prices dipped on uncertainty around President Trump’s proposed coalition to protect tanker traffic in the Strait of Hormuz amid the Iran war, before crude turned higher this morning. Amazon expanded three-hour delivery to 2,000 U.S. cities (one-hour in hundreds) with a Prime discount of $10 on fees. Apollo’s John Zito warned private equity may be overstating software valuations, flagging potential valuation risk for private credit and software loans.

Analysis

Nvidia’s product acceleration is shifting risk from single-vendor procurement to capacity and integration bottlenecks: customers will increasingly buy turnkey stacks (compute + software + services) which favors platform integrators and hyperscalers while compressing margins at smaller system integrators and some legacy ASIC suppliers. Expect a multi-quarter cadence where revenue recognition clusters around platform qualification windows, producing lumpy but higher-margin bookings; this amplifies earnings volatility even as aggregate TAM expands.

The self-driving and orbital compute pushes create non-linear adjacencies — automotive Tier-1s face potential margin compression and SKU rationalization as software-defined stacks replace bespoke silicon, while satellite and launch suppliers see demand for space-qualified compute and thermal/power solutions. Those secondary markets (space-grade power, rad-hard memory, high-reliability thermal management) are small today but can grow to represent high-margin annuity streams over 3–5 years.

Amazon’s faster-delivery scale changes unit economics in predictable ways: pickup of order frequency and willingness-to-pay will boost average revenue per user, but capital and labor intensity of sub 3-hour fulfillment will keep incremental margins muted until automation density rises. That implies a multi-year rollout where local real estate, robotics vendors, and short-haul logistics become the operating lever — winners will be those that monetize density (fees + subscription) rather than simply absorb cost to chase share.

Geopolitical shipping risk remains an asymmetric tail: insurance and freight-cost shocks can erupt quickly and add 5–15% to energy and logistics inputs over weeks, forcing margin re-pricing across retail and industrial supply chains. Separately, the private-markets valuation disconnect in software credit is a latent catalyst for markdowns that could ripple into CLO and bank trading desks if macro slows, with effects concentrated over the next 3–9 months.

Key Decisions for Investors

• Long NVDA via a 3-month call spread (buy 1 ATM call, sell 1 20–30% OTM call) to capture near-term product monetization while capping cost; target 2:1 reward:risk if GTC follow-through and order flows sustain, stop/hedge on a 12% drawdown in the underlying within 6 weeks.
• Pair trade: Long AMZN (90–180 days) / Short UPS or FDX (90–180 days) — capture last-mile share capture and monetization at Amazon vs margin pressure at incumbent carriers. Position size 2–3% NAV, target 30–50% relative outperformance, unwind if AMZN’s same-day fees are shown to materially compress take-rates across categories.
• Tactical overweight (6–18 months) in semiconductor capital equipment and high-reliability components (e.g., ASML, LRCX, or specific smallcaps in space-grade power/memory) to play secondary demand from datacenter + orbital compute. Risk: order push-outs; take profits on a 30–40% move or if reported customer cancellations exceed 10% of backlog.
• Event hedge: Buy short-dated (1–3 month) crude/oil call spread or XLE call spread as asymmetric protection against a sudden spike in shipping/energy risk from the Strait of Hormuz or other geopolitical flare-ups. Keep allocation small (<1% NAV) as insurance — payoff should offset 5–10% market-wide margin compression scenarios.