Stocks making the biggest moves premarket: Micron, Sandisk, Newmont and more

Google's new AI model prompted memory chipmakers to tumble (SanDisk -4%, Micron/Western Digital/Seagate ~-2%) after it could reduce memory needs for LLMs. Navan jumped 18% after guiding FY2027 revenue of $866–874M vs FactSet $840.8M, while multiple downgrades hit Qualcomm (-~2%) and Adobe (-1.4%). Mining names fell with metals (First Majestic -5%, Coeur/Hecla ~-4%, Newmont/Freeport ~-3%), and individual companies posted sharp moves: MillerKnoll -17.5% after Q3 adj. EPS $0.43 on $926.6M revenue and a $8–9M Q4 Middle East impact, Worthington Steel -13% on weaker Q3 EPS $0.27, and Snap -1.3% on an EU probe.

Analysis

Google's memory-efficiency step function is a demand reallocation event, not a literal overnight removal of addressable TAM. Expect an inventory digestion phase across DRAM/NAND over the next 2–9 months as OEMs and hyperscalers re-optimise BOMs and push spot pricing to clear legacy designs; that will pressure incumbent suppliers' utilization and force near-term capex deferrals that amplify price declines beyond the initial sell-off.

Second-order winners include suppliers of compute accelerators and software stacks that monetise model efficiency (licensing, managed inference), while hardware vendors heavily weighted to external memory content are structurally exposed. The impact will bifurcate within semis: firms with diversified end-markets or IP-driven revenue streams should see lower cyclicality, whereas commodity memory and high-logistics furniture names will face margin compression from both price and cost-side shocks.

Commodity-sensitive miners will show pronounced cash-flow swing risk over the next 1–4 quarters driven by spot metal volatility and hedgebook rolloffs; low-cost producers with minimal rolling hedges are the safest carriers of a prolonged metals recovery, while silver-heavy balance sheets carry higher downside. Regulatory and geopolitical noise (EU child-safety probe dynamics, Middle East shipping interruptions) magnify short-term volatility and could create unnatural dislocations that are tradable but must be time-boxed to event windows.

Key Decisions for Investors

• Pair trade (6–9 months): short a basket of memory-exposed names (MU, WDC, SNDK) via put spreads sized to limit max loss to premium paid; offset with a modest long in GOOGL (buy 6–12 month call spread). Rationale: asymmetric downside in memory suppliers during inventory digestion vs optional upside in software/cloud exposure. Target 30–50% downside on the short leg with a capped loss (premium); stop if GOOGL underperforms the market by >15% over 6 weeks.
• Mean-reversion long (3–6 months): if any memory name gaps >8% on headline-driven flows, layer into 3-month call debit spreads (buy low-cost calls and sell higher strikes) to capture bounce from buybacks/earnings positioning. Risk/reward: pay <100–150 bps of notional for 20–40% upside, stop at 12% premium loss.
• Event-driven long (12 months): accumulate NAVN on pullbacks up to 10% from intra-day highs or buy 12-month call spreads to express revenue-guidance durability. Risk: execution misses/cost inflation; reward: 30–60% upside if corporate travel momentum sustains. Place 10–15% stop-loss on equity positions.
• Relative-value metals trade (3–6 months): long a low-cost gold producer (NEM) and short a silver-centric small producer (AG) to isolate metal exposure and hedge macro. Rationale: gold is a clearer macro hedge vs silver's industrial cyclicality; target 20–35% net return, stop if spot gold moves >7% against the pair within 7 trading days.