CNBC Daily Open: Does Nvidia need a hug?

Nvidia set a high bar with a forecast-beating projection for fiscal 2028 sales to soar 70%, with CEO Jensen Huang saying demand is greater than 70% and supply supports delivering that pace. Reports also say Nvidia is in talks to buy Hugging Face for about $13 billion, while Anthropic agreed to a roughly $45 billion cloud deal to rent ~460MW of compute from Nscale. In related market noise, Meta reached a $16.7 billion settlement over alleged misstatements about child mental health harms, and oil stayed depressed after Iran said it reached an Oman deal to resume Strait of Hormuz vessel traffic.

Analysis

The real message is not “Nvidia is growing” but that AI demand is now outrunning the market’s ability to translate it into revenue, which supports pricing power and keeps the entire AI supply chain on allocation. That is bullish for NVDA’s multiple in the next 1-3 months, but the second-order winners are the bottlenecks: advanced packaging, HBM, networking, and data-center buildout. Hyperscalers such as AMZN, GOOGL, and META face a near-term free-cash-flow tax because they must keep capex elevated to avoid falling behind, even if monetization lags.

The M&A angle matters more as a strategic signal than as an EPS event. If NVDA moves further up the stack into developer distribution and model adjacency, it can deepen ecosystem lock-in, but it also raises antitrust and integration risk over 6-18 months. The bigger contrarian point is that the market may be treating supply-constrained growth as fully monetizable; if capacity ramps slower than expected, headline demand can stay huge while revenue upside disappoints.

META looks like the cleaner relative loser: the settlement is manageable in absolute dollars, but the more important effect is added friction on a business already exposed to engagement-quality scrutiny and teen usage constraints. That creates a slower path to multiple re-rating versus NVDA, and it could compress ad-monetization optionality if product changes reduce time spent. Falsifiers are simple: NVDA guide needs to keep exceeding consensus by enough to justify capex stickiness, and META needs to show no measurable engagement or ad-load hit within 1-2 quarters.

Key Decisions for Investors

• Long NVDA on post-print consolidation, preferably via 3-6 month call spreads rather than chasing spot; thesis is continued multiple support from supply-constrained growth. Exit or cut if the next guide stops beating consensus materially or gross margin/lead-time commentary suggests demand is normalizing faster than supply expands.
• Pair trade: long NVDA / short META over the next 1-3 months. NVDA has direct monetizable scarcity; META carries fresh regulatory/product-design drag with less obvious upside from the current AI cycle. Falsify the short if META engagement and ad pricing remain stable despite the settlement and product constraints.
• Watchlist, not immediate trade: AMZN and GOOGL as hidden losers from an AI capex arms race. If their next cloud-capex commentary re-accelerates without commensurate revenue lift, expect FCF margin pressure and multiple compression over 1-2 quarters; if monetization catches up, stay neutral.
• Do not chase the Hugging Face rumor alone. Only treat NVDA up-stack M&A as actionable if a deal is confirmed and terms imply strategic control of developer workflow; otherwise it is narrative support, not a standalone earnings catalyst.