Here are Thursday's biggest analyst calls: SpaceX, Nvidia, Tesla, Intel, Micron, Oracle, Meta, AMD & more

Wall Street saw a broadly positive set of ratings actions, including new Buy calls on SpaceX, Sensei Therapeutics, Septerna, Voyager Technologies, Advanced Energy Industries, and upgrades on Intel, Henry Schein, Black Hills, General Dynamics, Arthur J. Gallagher and CME Group. Bank of America also reiterated Buy on Oracle, Meta and AMD, while Wolfe raised Micron's price target to $1,250 from $550 and Oppenheimer reiterated Tesla as Perform. The only notable negative call was MoffettNathanson downgrading Chewy to Neutral from Buy.

Analysis

The common denominator is not “AI” but bottleneck ownership: compute, packaging, power, and distribution are the scarce layers that get repriced first when capex stays elevated. That favors semis and infrastructure-adjacent names with hard capacity constraints and visible backlog, while software-only beneficiaries remain more narrative-driven and vulnerable to dilution or capital intensity disappointment. Intel’s upgrade matters less as a standalone call and more as a signal that the market is beginning to price lead-edge wafer/packaging scarcity as a strategic asset, which could compress the valuation gap versus more cyclical peers if execution holds for 2-3 quarters.

The strongest second-order winner is the supply chain around AI buildout: memory, precision power, exchange/financing infrastructure, and utilities/transmission all gain from sustained data-center and capital-market activity. Oracle and Meta are being treated as funding vehicles for the AI cycle, but that creates a subtle risk: the more investors accept open-ended capex, the more the market will punish any deceleration in utilization or monetization. Micron’s raised target is the cleanest expression of the current regime — if HBM pricing stays tight through 2026, the earnings inflection can outrun consensus by a wide margin, but that also makes the trade increasingly crowded and vulnerable to even modest inventory normalization.

Outside tech, the more interesting setups are in regulated infrastructure and defense where duration is underappreciated: Black Hills, General Dynamics, and CME benefit from long-cycle visibility and less sensitivity to near-term macro noise. The contrarian angle is that the market may be overpaying for “AI adjacency” where the actual economic leverage is weak; names like Tesla and Chewy are more exposed to sentiment than fundamentals here, while SpaceX-related enthusiasm could pull capital away from public comparables before any hard monetization data exists. In biotech, the initiations are classic binary-upside setups, but runway buys only work if the next 2 readouts arrive cleanly; otherwise the discount rate rises fast and the PIPE cushion matters less than expected.

Key Decisions for Investors

• Long INTC / short a basket of slower-moving large-cap semis over 1-3 months; the relative trade is for supply-chain scarcity to re-rate Intel faster than consensus expects, with downside capped if execution disappoints because the short basket should still benefit from the same AI capex theme.
• Buy AMD on pullbacks and pair with a partial hedge in NVDA over the next 2-4 weeks; AMD has more catch-up optionality into the next product cycle, while NVDA remains the higher-quality franchise but is more exposed to any digestion in data-center spend.
• Initiate a 6-12 month long BKH vs short utility index basket; the merger/scale angle plus data-center and transmission exposure should outperform regulated utilities if infrastructure capex stays firm, with lower earnings risk than pure growth names.
• Use AEIS as a leveraged way to express the power-conversion bottleneck for AI buildout; best entry is on post-event weakness, targeting 15-20% upside over 3-6 months if demand visibility persists.
• Avoid chasing CHWY on weakness; if holding, reduce into any strength and rotate into higher-conviction AI-adjacent industrials, because the current issue is not valuation but slowing top-line momentum with limited catalyst density.