The Dow's rough month, Micron's revenue surge, Google launches Gemini 4 and more in Morning Squawk

Micron beat fiscal Q4 expectations and guided above consensus, with data-center revenue rising 11-fold year over year, though shares edged lower premarket after a more than 500% 12-month rally. Markets entered Q4 with weak momentum: the Dow fell more than 4% in September, while the 10-year Treasury yield rose 4bps overnight to its highest level since 2002 as Fed official Neel Kashkari said inflation remains near 3% and still too high. Alphabet began limited rollout of Gemini 4 Argon, while Eli Lilly reported superior mid-stage weight-loss results for an eloralintide-tirzepatide combination versus tirzepatide alone.

Analysis

MU’s setup has shifted from an AI-demand validation story to an expectations-and-duration risk. After a parabolic rerating, even guidance upside can fail to sustain the stock if investors infer that hyperscaler inventory is being pulled forward or if DRAM/NAND pricing peaks sooner than expected; the key 1-3 month read-through is HBM allocation, contract-price commentary, and gross-margin trajectory rather than reported revenue. The cleaner second-order beneficiaries remain equipment and materials suppliers with multi-quarter capacity commitments (LRCX, KLAC, AMAT), while PC- and handset-exposed memory demand remains the weak link.

GOOG’s internal use of a stronger model for data-center efficiency matters more economically than benchmark leadership: lower inference/training cost per workload can defend Cloud margins while allowing price competition against MSFT and AMZN. The near-term revenue impact is unprovable, and enterprise cyber adoption cycles are long, but a credible efficiency advantage can reduce the market’s concern that Alphabet’s AI capex is structurally dilutive over the next 6-18 months. This is modestly negative at the margin for standalone cybersecurity vendors only if Google bundles AI security functionality into existing Cloud contracts; it is not yet a reason to short PANW or CRWD.

Rising long-end yields create a valuation headwind for high-multiple AI and obesity franchises even where fundamentals remain intact. LLY’s combination data expands the addressable pool of inadequate responders, but phase-three design changes introduce dose, tolerability, and discontinuation risk; this is a 12-24 month pipeline option rather than a near-term earnings catalyst. The market is likely underpricing the possibility that payer utilization controls, not efficacy, become the binding constraint on incremental GLP-1 revenue.

The practical near-term regime is selective earnings ownership rather than broad index beta: META and GOOG have cash-rich balance sheets and self-funded AI spend, whereas long-duration semis are more exposed if real yields continue rising. A sustained move higher in yields without upward EPS revisions would favor quality mega-cap platforms over the AI hardware complex.

Key Decisions for Investors

• Maintain MU only as a tactical 1-3 month long after post-earnings volatility settles; use a 8-10% downside stop from entry and reduce if next-quarter gross-margin guidance fails to rise sequentially. Upside requires evidence that HBM supply remains constrained into the next fiscal year; absent that, risk/reward is unfavorable after the rerating.
• Initiate a 3-6 month pair: long GOOG / short an equal-dollar basket of SMH. The thesis is that Alphabet can monetize and internally harvest AI productivity with lower balance-sheet and yield sensitivity than the semiconductor complex; exit if GOOG Cloud margin compresses or 10-year yields reverse sharply lower, restoring hardware-duration leadership.
• Treat LLY as a core long only on 6-18 month pullbacks rather than chasing trial headlines. Add if the stock retraces 10-15% without a reduction in obesity-volume guidance; invalidate on materially higher discontinuation rates in phase three or adverse payer-access commentary.
• Avoid a directional AAPL trade based on political access optics. The actionable watch item is whether exclusion is followed by concrete procurement, antitrust, tariff, or AI-policy actions; without that, it has no durable earnings transmission mechanism.
• For portfolio hedging over the next 1-3 months, favor trimming high-multiple AI hardware exposure rather than shorting broad equities if the 10-year yield makes new highs. A reversal in core inflation or a dovish Fed shift would falsify the rate-pressure thesis and argues for covering the hedge.