Stocks making the biggest moves midday: Marvell, Moderna, Lamb Weston, Constellation Energy & more

Trading was mixed: Vistra rose 12% after a $4.2 billion federal nuclear investment announcement, while Constellation Energy gained 14% on a Google power agreement linked to 890 MW of new nuclear capacity and a separate 2,700-MW supply deal. Marvell projected fiscal 2028 revenue of $20 billion versus $18.17 billion consensus and fiscal 2031 revenue of $70 billion–$90 billion versus $47.04 billion consensus; Lamb Weston beat quarterly estimates and raised its full-year adjusted EPS outlook to $3.05–$3.35. Option Care Health surged more than 30% on a report of advanced acquisition talks valued above $5 billion, while memory-chip shares fell as investors reassessed data-center demand and profit prospects.

Analysis

The tape is separating AI infrastructure exposure by perceived bottleneck, not confirming a uniform change in AI demand. Storage weakness alongside strength in networking and custom silicon may reflect rotation toward compute/connectivity; it is not yet evidence that data-center investment is being cut. For Seagate (STX), Western Digital (WDC) and SanDisk (SNDK), the key test is whether subsequent customer capex commentary or company guidance actually lowers storage demand, rather than a one-session sentiment reset. Also distinguish HDD exposure at STX/WDC from SanDisk’s flash exposure.

Marvell’s long-dated outlook creates asymmetric execution risk: FY2028 expectations are nearer-term testable, while the FY2031 range is especially sensitive to design-win conversion, customer concentration and deployment timing. Treat it as a forecast, not contracted revenue. Likewise, the CPU-demand thesis for AMD depends on AI agents translating into incremental server CPU deployments rather than displacing or merely reshuffling existing workloads.

Power headlines support a structural case for firm generation, but not an immediate earnings uplift: nuclear capacity additions require long development and delivery timelines, and announced investment is not equivalent to approved, financed capacity or attractive realized power prices. Constellation’s customer-specific agreement is more directly tied to demand visibility than a broad policy announcement; verify contract economics and milestones before capitalizing it. Near term, the largest reversal risks are weaker hyperscaler capex, delays in power projects, and guidance that fails to validate optimistic AI forecasts. The Option Care report is deal-binary; absent a signed agreement and terms, the share move may price more certainty than the report provides.

Key Decisions for Investors

• Avoid chasing the AI infrastructure rally at the open. Consider a staged relative-value position long Ciena (CIEN) versus a basket of STX/WDC only after confirming CIEN order/backlog strength and no storage-demand downgrades; exit if networking orders soften or storage guidance stabilizes while CIEN expectations rise further. This is a rotation trade, not a claim that storage demand is structurally impaired.
• Keep Marvell (MRVL) and AMD on catalyst watch rather than underwriting headline targets. Add only if the next guidance cycle validates nearer-term revenue conversion and customer deployments; cut the thesis on design-win delays, capex reductions, or evidence that AI-agent workloads do not lift server CPU demand.
• Do not extrapolate nuclear announcements directly into near-term earnings. Track CEG contract economics, permitting, financing and delivery milestones; a delay or unattractive pricing would undermine the valuation premium. Treat VST’s policy-driven rally as more exposed to execution and policy reversal until project-level details are disclosed.
• Treat Option Care (OPCH) as event risk, not a confirmed acquisition. Avoid paying the deal premium before definitive terms; a report denial, financing/partner withdrawal, or materially lower consideration is the key downside trigger.