Nvidia CEO Jensen Huang eats dinner with someone ‘and their stock price doubles the next day’

Nvidia’s Q2 results surged investor sentiment: revenue rose to $96.2B vs $92.2B consensus (+70% revenue growth projected for next fiscal year), driving the stock up more than 4% after-hours. On the supply-constrained outlook, Jensen Huang said Nvidia currently has supply for ~70% of current customer demand and that the entire supply chain is “challenged,” with heavy constraints (e.g., data center power, memory pricing, shell availability, and wafer access). Despite the bottlenecks, the significantly higher growth figure and supply ramp visibility helped propel a bullish repricing.

Analysis

The real signal is not the dinner optics; it is that the AI supply chain is still the binding constraint, which usually pushes incremental economics upstream. That favors the scarcest physical inputs—advanced wafers, packaging, and memory—more than the brand names attached to the customer list, so TSM and SSNLF have a cleaner second-order read-through than ORCL, which still needs chips in hand before it can monetize capacity.

NVDA’s setup remains structurally strong, but the marginal upside from here is increasingly about execution against an already elevated bar, not just demand discovery. If supply remains the choke point over the next 1-3 months, the market should keep paying for scarce compute exposure; if capacity comes online faster than expected, the multiple can compress even while fundamentals stay good, because scarcity premium fades.

The contrarian mistake is treating the celebrity-customer halo as fundamental alpha. That effect is mostly a short-lived sentiment/liquidity trade in local affiliates and adjacent consumer names, not a durable earnings channel. The more durable risk is that prolonged chip scarcity accelerates hyperscaler custom-silicon efforts over 6-18 months, gradually eroding NVDA’s pricing power at the margin even as near-term revenue stays superb.

Key Decisions for Investors

• Stay long NVDA on pullbacks over the next 2-4 weeks; do not chase the after-hours gap. Falsifier: if next-quarter supply commentary or backlog conversion implies materially faster capacity relief, trim exposure.
• Prefer a long TSM position versus chasing ORCL’s AI monetization narrative over 1-3 months. The former captures scarce-capacity economics directly; the latter is still gated by GPU availability.
• Fade the event-driven halo in HYMLF and NHNCF after any opening pop; use a 1-5 day horizon and small size because the fundamental linkage is weak. Risk/reward skews to mean reversion once the narrative cools.
• For a cleaner hedge, use NVDA calls financed by reducing exposure to other AI-adjacent momentum names that have already rerated on sentiment rather than earnings. This keeps upside to the core winner while avoiding the weakest second-order beta.