Stocks making the biggest moves midday: Palo Alto Networks, Fervo Energy, Chevron, Apple & more

Construction and AI-linked infrastructure stocks sold off midday: U.S. construction spending fell 0.5% in July to the lowest since Oct 2023, with ITB down ~2%, Builders FirstSource -4% and LGI Homes -3%, while AI infrastructure/photonic names like Lumentum (-5%) and Teradyne (-5%) declined on a risk-off tape. In contrast, Moderna jumped 5% after late-stage cancer-vaccine results momentum, Novartis rose 6% on multiple-sclerosis trial data, and Medtronic gained 2% after raising FY2027 EPS guidance to $5.94–$6. Energy and semiconductors were mixed: XLE climbed ~1% on Middle East oil-tension support (oil-linked EOG +2%), while Treasury yield pressure weighed on chipmakers as SMH fell >2% and AMD slipped ~3%.

Analysis

The housing tape looks more like a rate-beta liquidation than a clean demand collapse, which matters for relative value. The most levered names tied to new construction and distributor sell-through should keep underperforming until mortgage rates roll over, while repair/remodel exposure is the better buffer; that argues for a continued spread trade between builders/materials and the more defensive home-improvement complex. If yields stay high for another 1-3 months, expect estimate cuts before the next housing print; if rates ease, this group can mean-revert quickly, so the signal is tactical rather than secular.

AI infrastructure is getting hit from a discount-rate shock, but the more important second-order effect is power scarcity. Platform owners with balance-sheet flexibility and electricity access can keep buying optionality, while capital-intensive suppliers of testing, photonics, and semiconductor equipment are the first place investors de-gross when long-duration growth is pressured. The Google/Fervo linkage is a real tell: the bottleneck is shifting from compute availability to 24/7 power procurement, which should widen the moat between hyperscalers and the rest of the AI stack over 6-18 months.

In healthcare, the market is rewarding headline momentum, but the dispersion is getting extreme and likely unstable. The biotech winner with the cleanest path is the one where positive data can still translate into a valuation rerate without requiring immediate commercialization assumptions; the names that already ran hardest are most vulnerable to a gap-down on any nuance. Cyber looks similar: pre-earnings weakness can be a positioning reset or an early warning, but the real catalyst is not the stock tape — it is whether management confirms demand durability and billings momentum.

Contrarian take: consensus is treating every weak tape as fundamental deterioration. In semis and cyber, some of this is just factor de-risking; in housing, the market may still be underpricing how long high rates suppress volumes. That asymmetry favors patience in growth beta and selectivity where balance-sheet strength and customer concentration actually matter.

Key Decisions for Investors

• Short BLDR / long HD for 1-3 months as a relative-value rates trade; thesis works if mortgage rates stay elevated and new-start volumes keep slipping, but should be cut if the 10Y meaningfully breaks lower and refi activity reaccelerates.
• Buy GOOGL on weakness as a hyperscaler-with-power-access winner; pair it against a basket short in rate-sensitive AI suppliers such as TER/COHR/LITE if you want to express the view that power and scale matter more than equipment beta over the next 6-18 months.
• Do not chase MRNA after the vertical move; if already long, trim into strength or express a small defined-risk short via call spreads, with the thesis invalidated by another clean positive oncology catalyst or partnership update.
• Stay neutral on PANW/CRWD until after earnings; if PANW merely meets but guides conservatively, use any post-print bounce to short the cyber basket, since the sector is vulnerable to multiple compression when growth visibility softens.