CNBC Daily Open: Iran, Oman in talks on Hormuz; SpaceX is loyal to Nvidia on AI

Markets pushed modestly higher overall with the Dow closing at a fifth straight record, but oil slipped on signs Iran and Oman are nearing an agreement that would give Tehran more control of Strait of Hormuz commercial shipping. The Fed outlook turned more hawkish as Fed Governor Lisa Cook said she’s “prepared to act” on a rate hike, with CME pricing a 55% chance of a September hike. Stock moves were conflicted: SpaceX shares fell 13.6% after AI capex jumped to $18.4B (sixfold) but AMD dropped 7% despite strong Q2 results and >100% data-center growth guidance following Musk’s claim SpaceX will use Nvidia exclusively; Alphabet slid ~4% as Google’s leadership changes coincide with capex forecast up to $205B.

Analysis

The market is treating the oil headline as a clean disinflation impulse, but the real mechanism is a lower near-term volatility regime: less crude tail risk means softer breakeven inflation, fewer upside surprises in CPI, and less pressure for the Fed to validate the hawks. That is bullish for long-duration assets over the next 1-3 months, but only if the Strait agreement proves durable; any shipping disruption would snap risk assets back into an inflation scare within days.

The AI tape is turning from “who can grow fastest?” to “who has procurement power and financing capacity?” A single strategic buyer signaling exclusive vendor preference is a stronger commercial signal than a quarterly beat, which is why NVDA deserves the premium while AMD risks being stuck as the alternative channel, even if its data-center growth stays strong. The bigger overhang is GOOGL: capex intensity plus negative free cash flow can compress the multiple even if core products remain healthy, because the market is now rewarding AI share capture more than broad ecosystem quality.

Contrarian view: the AMD selloff may be too reflexive if this is one customer’s bias rather than an industry-wide replacement cycle, and GOOGL may not deserve a structural short if management can slow spend after the current buildout. The consensus may also be underestimating second-order winners from higher-for-longer rates: anything dependent on terminal value and cheap financing will stay hostage to the September Fed path, so a hawkish data surprise could unwind the entire AI-beta complex faster than fundamentals change.

Key Decisions for Investors

• Long NVDA / short AMD as a 1-3 month relative-value trade; entry on any AMD rebound toward the post-earnings gap. Thesis: procurement concentration should widen valuation dispersion even if both grow. Falsify if AMD’s next data-center commentary shows new hyperscaler wins or NVDA fails to translate share gain into upside guide.
• Long META / short GOOGL into the next earnings cycle; the better risk/reward is on the company that can fund AI with existing cash generation rather than one where capex is visibly eroding free cash flow. Falsify if GOOGL signals capex moderation or meaningful monetization improvement faster than expected.
• Avoid chasing outright short AMD or outright short GOOGL here; use pairs instead. The market is already pricing a lot of the disappointment, and direction-only shorts are vulnerable to any broad risk rally if crude stays soft and Fed hike odds fade.
• Set a tactical alert on Brent and the September FedWatch probability: if crude reverses higher and hike odds move above ~65%, reduce exposure to high-multiple AI names and add hedges. If both reverse lower, the current risk-on rotation can extend for several weeks.