Is Micron Stock a Buy Now?

Micron shares have surged ~180% over the past six months as AI-driven demand has fully consumed memory capacity and driven chip prices and profits higher. The HBM TAM is projected to grow from $35B in 2025 to $100B by 2028, and Micron expects its Idaho fab online mid-2027 with a second facility by 2028. The market is pricing in cyclicality — the stock trades at ~11x forward EPS — so a supply rebound could trigger a sharp price and profit correction; the position requires active monitoring.

Analysis

Tight HBM availability has created a temporary mismatch where a commoditized product behaves like a differentiated bottleneck — that changes bargaining dynamics with hyperscalers and OEMs and turns short-term utilization into a de facto pricing lever. Because memory fabs are capital- and time-intensive, marginal dollar flows today disproportionately accrete to suppliers who are already at scale; that creates a multi-year window where FCF and free-cash conversion can diverge widely across players even if unit costs look similar.

Second-order winners include capital-equipment and substrate vendors whose order backlogs extend lead times for any fast-follow capacity; conversely, ODMs and smaller cloud players who cannot secure long-term HBM contracts face either margin erosion or accelerated capex to vertically integrate. Architectural shifts (CXL disaggregation, weight quantization, sparsity, on-die SRAM for inference) are plausible demand dampeners — each reduces HBM/GPU attachment ratios and can materially shorten the cycle if adopted by a few hyperscalers simultaneously.

Key timing risks are not just calendar (fab ramp dates) but procurement cadence: hyperscaler buying often comes in multi-quarter tranches that can cause step-function revenue moves; likewise, a pair of aggressive capex announcements from Samsung/SK could flip supply into oversupply within 12–24 months. That makes this a trade that should be managed around discrete catalysts (fab ramp milestones, quarterly contract wins, HBM spot indices) rather than treated as a buy-and-forget secular AI play.

Key Decisions for Investors

• Long-duration, hedged MU LEAPS: Buy MU Jan-2028 calls (size ~1.5–2% NAV) to capture multi-year HBM tightness; fund by selling near-term (3–6 month) MU calls to monetize IV. R/R: asymmetric upside if scarcity persists; downside limited to net premium. Review at Idaho ramp milestone and again at Jan-2028.
• Relative-value pair (capitalize on memory-cycle reversion): Long MU equity / short NVDA equity sized to neutralize market beta (initially target 0.6 NVDA shares per MU share). Timeframe 3–12 months; target 20–30% relative reversion in MU/NVDA spread. Stop-loss: 15% on combined position or if NVDA reports demand shocks that materially increase GPU attachment rates.
• Tactical protection and signals: Buy 3–6 month MU puts sized to LEAP position (cost ~0.25–0.5% NAV) ahead of key supply announcements; alternatively sell MU near-term covered calls to generate income if comfortable with capping upside. Set automated alerts for (a) Samsung/SK hynix capex announcements, (b) weekly HBM spot-index moves ±10%, and (c) hyperscaler contract disclosures — any of these should trigger re‑weighting within 48–72 hours.