Top 10 things to watch in the stock market Friday

Akamai secured a seven-year, $11.6 billion expanded compute commitment from Anthropic, sending its shares nearly 20% higher and reinforcing bullish expectations for AI-driven CPU demand across Intel and AMD. AMD's Bank of America price target was raised to $720 from $620, while BMO said Micron's memory market remains in a supply-constrained supercycle through 2027. Macro conditions remain mixed: the 30-year mortgage rate has climbed to 7.45%, its highest since 2024, while oil fell on reported prospects for Iran to reopen the Strait of Hormuz and resume U.S. nuclear talks. Retail signals were uneven, with Costco beating Q4 expectations but retaining membership-growth concerns, while Bank of America downgraded Nike to sell and cut its target to $30 from $47.

Analysis

AKAM’s gap should not be extrapolated one-for-one into earnings: a multi-year compute commitment can be heavily pass-through infrastructure spend, with revenue recognition, gross margin, minimum-volume guarantees, and customer concentration still undisclosed. The more durable read-through is that agentic workloads broaden AI demand beyond GPU training toward latency-sensitive inference and general-purpose compute, improving the strategic case for INTC and AMD; however, AMD’s valuation now leaves little room for a datacenter execution miss. Over the next 1-3 months, contract-economics disclosure and hyperscaler capex commentary matter more than the headline commitment.

MU is the cleaner near-term AI hardware catalyst because memory pricing and utilization can translate into earnings faster than CPU-share narratives, but the supercycle thesis requires disciplined supply behavior through 2027. A single weak pricing or inventory signal at results could unwind high-expectation semiconductor positioning broadly, including AMD, even if long-run AI demand remains intact. Favor MU exposure into earnings only with defined downside; use post-results guidance on HBM/DRAM pricing, bit supply, and customer inventory as the thesis test.

Higher long-end yields create a less obvious dispersion trade: trust banks such as BNY can retain asset-servicing and net-interest-income support without requiring a strong underwriting calendar, while GS is more exposed to capital-markets activity and risk appetite. Conversely, sustained mortgage-rate pressure is a delayed headwind for discretionary consumer demand and housing-linked spending; COST’s membership model is relatively defensive, but its premium multiple leaves it vulnerable if renewal or comparable-sales trends soften. A de-escalation-driven oil decline could ease inflation expectations and eventually help duration-sensitive equities, but only if it translates into lower real yields rather than merely lower energy prices.

Key Decisions for Investors

• Do not chase AKAM’s opening move; place a 1-2 week watch on disclosed contract structure, implied annual revenue contribution, gross-margin treatment, and customer termination protections. Consider a tactical long only if the stock retains at least half of the initial gain after those details emerge; falsify on evidence that the commitment is primarily low-margin pass-through capacity.
• Express AI compute broadening via a 3-6 month long INTC / short AMD pair, sized modestly. INTC has more upside torque if enterprise CPU inference demand improves, while AMD is more vulnerable to multiple compression after a near-vertical move; exit if AMD delivers material datacenter-share gains or INTC cuts foundry/client guidance.
• Use MU earnings as a defined-risk event: buy a small call spread 1-3 months out rather than unhedged shares, contingent on options implying a move below prior earnings reactions. Take profits on confirmation of improving HBM and DRAM pricing; cut if management signals rising customer inventories, supply additions, or weaker calendar-2027 pricing.
• Initiate a 1-3 month long BNY / short GS relative-value position if the 10-year yield remains elevated and equity-issuance/M&A activity does not accelerate. The trade is invalidated by a sharp rally in capital-markets volumes, a meaningful drop in yields that compresses BNY’s rate benefit, or a GS upside surprise in advisory and trading.
• Maintain a bearish bias on NKE but avoid initiating after a downgrade-driven air pocket; sell rallies over the next 1-3 months unless evidence emerges of improving full-price sell-through and gross-margin recovery. Cover on a credible upward revision to revenue or margin guidance, which would indicate the turnaround timeline is shortening.