Micron and General Mills have been highlighted as Zacks Bull and Bear of the Day

Micron (MU) is highlighted as a Zacks Rank #1 bull on AI data-center demand for HBM, with claims of a 13th consecutive Zacks consensus EPS beat and 2026 EPS expected to grow nearly 800%. General Mills (GIS) is flagged as a Zacks Rank #5 bear as GLP-1 weight-loss drugs suppress appetite, with management projecting fiscal 2027 category growth below long-term trends and consensus EPS declining through 2028, then rising only ~4.73%. GE Vernova (GEV) is covered ahead of July 22 results, with 2Q 2026 consensus EPS of $3.17 and revenues of $10.77B (up 18.2% YoY), plus expectations for ~15%–17% power revenue growth and EBITDA margin expansion, partially offset by offshore wind EBITDA losses of $200M–$300M.

Analysis

MU is the cleanest way to express a second-order AI bottleneck trade: when memory, not compute, is the constraint, pricing power migrates upstream and the winners are the few suppliers with qualified HBM capacity. That favors MU, but it also raises the bill of materials for NVDA’s ecosystem and could eventually slow server deployment if memory inflation outruns customer budgets. Near term, the market should still reward any evidence of sold-out capacity and contract pricing; the larger risk is a 6-18 month overbuild cycle if supply catches up faster than AI demand.

GIS sits on the wrong side of a slow-burn volume erosion story. The immediate issue is not one quarter of weaker traffic; it’s the compounding effect of smaller baskets, heavier promotions, and trade-down to private label as consumers remain price-sensitive. The contrarian nuance is that GLP-1 adoption is real but its direct sales impact is probably more gradual than the bearish narrative implies; the more durable pressure is margin dilution from defending share, not a sudden demand cliff. NVO and LLY are the structural beneficiaries of that redistribution.

GEV is a better fundamental story than a clean valuation story. AI-driven power demand and grid capex are creating a multi-quarter backlog tailwind, but at a premium multiple the stock is vulnerable if orders normalize or if wind losses remain sticky longer than expected. The key catalyst window is the next earnings cycle: a clean beat can extend momentum, but any guide-down on margin or backlog conversion would likely trigger multiple compression quickly.

Key Decisions for Investors

• Long MU on pullbacks over the next 1-3 weeks; highest-conviction expression of AI capex scarcity. Falsify if HBM pricing softens, qualification timelines slip, or customer inventory commentary turns cautious.
• Initiate a 3-6 month short GIS position or buy put spreads; the best risk/reward is on margin compression from promo intensity and private-label substitution rather than top-line collapse. Cover if organic sales re-accelerate or management meaningfully lifts forward guidance.
• Pair trade: long NVO / short GIS for 3-6 months to isolate GLP-1 substitution versus packaged-food demand erosion. Risk is reimbursement friction or slower-than-expected GLP-1 adoption.
• Do not chase GEV after strength; use any post-earnings pop to trim or hedge with call spreads because the premium valuation leaves little room for execution misses.
• Watch NVDA supplier-cost commentary into the next AI hardware cycle; if memory inflation starts to pressure server pricing, the trade may rotate from compute leaders to component suppliers like MU.