Here are Thursday's biggest analyst calls: Nvidia, Apple, Tesla, CoreWeave, Meta, Dick's, ResMed & more

Wall Street saw a wave of rating actions, with several bullish moves including JPMorgan initiating ResMed at overweight, Evercore reiterating Nvidia as outperform, and upgrades on Okta, Hamilton Lane, Blackstone Mortgage, Flywire, BP, and Descartes. Notable negatives included Citi downgrading Flutter to sell, Mizuho cutting PayPal to neutral, and JPMorgan downgrading Qualcomm to neutral amid rising competition from Nvidia. The set of calls is likely to drive individual stock moves, but the overall readthrough is mixed to modestly positive for AI, software, and financial-services names.

Analysis

The tape is rewarding businesses with visible operating leverage and punishing names where the market sees structural share loss. The clearest second-order winner is the AI infrastructure complex: if enterprise AI spend is truly shifting from pilots to production, the beneficiaries extend beyond model vendors into identity, networking, and capacity-financing layers, which supports OKTA, CRWV, and META more than the consensus implies. That same lens is a headwind for QCOM and PYPL: both face an increasingly crowded distribution stack where platform owners can compress margins by bundling adjacent functionality.

The near-term setup is strongest in names where analyst upgrades coincide with a plausible catalyst over the next 1-2 quarters. RMD, BXMT, HLNE, and FLYW are less about multiple expansion and more about duration: each has a path to re-rate if the market gains confidence in recurring earnings power and free cash flow durability. By contrast, C and PYPL still read as fragile because the market is likely to demand evidence of clean inflection before rewarding them; negative revisions can persist for several quarters even after valuation looks cheap.

Contrarian take: the market may be underestimating how much of the AI-benefit pool gets captured by picks-and-shovels rather than the obvious mega-caps. That argues for being selective on NVDA and AAPL here—both remain structurally well positioned, but the easier money may come from second-order enablers whose earnings revisions can surprise more sharply off a lower base. The biggest risk to the bullish stack is a short-cycle digestion period in AI capex or a post-earnings reset in the consumer/auto names, which would keep the market in a rotation rather than a broad risk-on move.

Key Decisions for Investors

• Long OKTA vs short QCOM for 1-2 quarters: pair captures the shift from device-centric compute to identity/security spend tied to enterprise AI rollout; risk is if AI adoption stalls and both de-rate together.
• Buy CRWV on pullbacks for a 3-6 month catalyst trade: differentiated financing capacity should matter as hyperscalers and enterprise customers keep adding capacity; use tight risk controls given execution/financing sensitivity.
• Prefer RMD and BXMT as defensive growth longs over high-beta software: both have clearer earnings durability and should hold up if the market rotates out of crowded AI winners.
• Avoid chasing PYPL into analyst negativity; if owning fintech beta, consider EFX/FICO instead where pricing power is more defensible and catalyst quality is higher.
• Tactically own NVDA, but hedge with a basket short in QCOM/PYPL: if the market is right that AI demand is broadening, NVDA should still lead; if not, the hedge should cushion a correction in adjacent tech.