Here are Wednesday's biggest analyst calls: Nvidia, SpaceX, Broadcom, SolarEdge, Nike, Oracle & more

Wall Street calls were broadly constructive, with multiple reiterations/overrides turning names into buys or upgrades: Morgan Stanley highlighted Williams Companies (WMB) as a top pick for AI infrastructure exposure, while Citi reiterated Oracle as a buy and BoA reiterated Nvidia at a $350 target with focus on balance-sheet and supply commitments ahead of earnings. On the solar side, UBS upgraded SolarEdge to buy, citing the FCC ban on new inverter model imports as a potential supply-constrained driver for pricing power. Offsets included Truist downgrades of Dick’s and Nike to Hold (PTs cut to $135 and $42), with both seeing “category pressures” and murkier turnaround signals.

Analysis

The cleanest signal is that AI capex is broadening from chips into infrastructure toll-takers. That favors contracted midstream and power-adjacent names like WMB over pure software multiple stories, because the cash flow is tied to physical buildout rather than sentiment around model adoption. By contrast, ORCL and NVDA remain higher-beta expressions of the same theme: both can work, but the market is increasingly pricing in multi-year commitments and balance-sheet support, so any disappointment on order quality, financing intensity, or margin mix could compress multiples fast.

On solar, the opportunity is more tactical than structural for SEDG. Import restrictions can create a short-lived pricing umbrella, but if the channel is still inventory-heavy the benefit may show up first in gross margin before translating into durable share gains; FSLR is the cleaner 6-18 month policy beneficiary because domestic supply is harder to displace. The risk is that investors overpay for the nearest-term beneficiary while underappreciating how quickly parallel sourcing, stockpiles, or customer substitution can mute the upside.

The consumer downgrades read like a demand-quality warning, not just a one-quarter miss. Weakness at discretionary and off-price tends to bleed into footwear/apparel suppliers, discount logistics, and mall traffic with a lag of 1-2 quarters, so NKE and TJX may stay range-bound until channel data stabilizes. If inflation cools but volumes do not, the margin reset can last longer than consensus expects, which argues for patience on any dip-buying.

Key Decisions for Investors

• Long WMB vs short ORCL into the next 1-3 months: prefer the contracted infrastructure cash flow over a software re-rating that now depends on proving multi-year monetization; thesis fails if ORCL guides materially above current spend expectations.
• Buy FSLR on pullbacks and treat SEDG as a tactical trade only: FSLR has the better 6-18 month durability if policy support persists, while SEDG upside is more vulnerable to inventory normalization; falsifier is U.S. distributor pricing staying firm for multiple quarters.
• Avoid chasing NVDA into earnings; use the print to assess off-balance-sheet commitments and gross margin durability before adding risk. If disclosures show commitments outpacing incremental revenue visibility, expect multiple compression over days to weeks.
• Short NKE on any bounce over the next 2-6 weeks, or pair against a more resilient apparel/brand basket if available: the risk/reward is favorable while category pressure is still being revised into estimates; cover if management cuts the turnaround timetable.
• Watch TJX rather than trade aggressively: if next-month comp trends do not re-accelerate, the stock can remain range-bound for months, but a clear traffic inflection would invalidate the short-consumer read.