Our top 10 things to watch in the stock market Thursday

Brent crude rose 2% to $105/bbl and the 30-year Treasury yield hit a post-2004 high of 5.446%, driving Nasdaq futures about 1% lower as higher energy costs and rates pressure equities. U.S.-China trade-truce extension through Jan. 10 and Trump-Xi talks may create catalysts for Boeing, while AI product and infrastructure developments supported Meta, Amazon, Alphabet and GE Vernova. Oracle fell more than 4% on a potential Project Jupiter payment delay, Starbucks will close 250 North American cafes and take $300M in restructuring charges, and Darden dropped over 3% after below-consensus full-year EPS guidance.

Analysis

The relevant macro transmission is not simply risk-off: a higher long-end discount rate combined with energy inflation pressures both long-duration AI valuations and discretionary demand. META, AMZN, and GOOGL can absorb this through cash flow, but their multiples are more vulnerable than near-term earnings suggest if real yields remain elevated for several weeks. SBUX and DRI face a less favorable setup: fuel-driven pressure on lower-income traffic arrives before menu-price actions can offset it, while additional store rationalization signals that fixed-cost deleveraging remains a risk.

ORCL's contractual contingency introduces a more important question than a single project delay: whether hyperscaler/data-center demand is becoming constrained by power, construction completion, and financing rather than GPU availability. This is incrementally negative for ORCL's backlog-to-revenue conversion and for OWL if project-finance counterparties seek revised terms; it is comparatively constructive for GEV, whose turbine backlog and pricing power benefit when grid interconnection delays force behind-the-meter generation. The thesis breaks if data-center developers curtail capex rather than substitute toward self-generation.

A China aircraft announcement would be sentiment-positive for BA, but its earnings value is capped until delivery cadence, supplier quality, and working-capital normalization prove durable. Treat any trade-policy headline as an option on order-book confidence, not a change in free-cash-flow estimates. Conversely, agentic commerce is likely to favor platforms controlling identity, discovery, payments, and fulfillment; META's upside depends on whether agents improve ad conversion rather than merely raise inference and hardware costs. Over the next 6-18 months, AMZN is better positioned to monetize closed-loop transaction data, while open-web retailers and parcel carriers face incremental price competition in low-density merchant shipments.

Key Decisions for Investors

• Initiate a 1-3 month pair: long GEV / short ORCL in equal dollar beta-adjusted sizing. The catalyst is evidence that power-constrained campuses shift spending toward gas generation while Oracle's committed-cloud revenue timing is reassessed; exit if ORCL reaffirms 2028 campus commissioning with binding funding terms or GEV reports turbine-order slippage.
• Reduce consumer-discretionary exposure via short DRI versus long COST through the next two earnings cycles. COST's membership model and grocery mix provide relative insulation from fuel-sensitive traffic; target 8-12% relative return, with a stop if DRI traffic materially accelerates or full-year margin guidance rises.
• Buy BA only on a confirmed China order coupled with monthly delivery-rate evidence, rather than ahead of diplomatic headlines. Use a 3-6 month call spread to cap event risk; falsify on another production-quality disruption or a failure to convert higher output into positive free cash flow.
• Maintain META exposure but fund it with a partial long AMZN hedge/pair rather than adding outright after the recent rerating. Over 6-18 months, AMZN has clearer transaction monetization from agentic shopping; reassess if META demonstrates sustained ad-conversion lift without a material increase in capex or operating-expense guidance.
• Set a macro risk trigger: if the 30-year Treasury yield holds above 5.5% for five sessions, cut net exposure to high-multiple AI/software positions and favor GEV, COST, and cash-generative mega-cap platforms. A rapid yield reversal driven by softer inflation or growth data would invalidate the duration-compression leg.