Back to News
Market Impact: 0.35

This Artificial Intelligence (AI) Stock, Up 28,700% Since Its IPO, Could Be the Biggest Bargain of the Decade

MUNVDAAMDAVGONFLXNDAQ
Artificial IntelligenceTechnology & InnovationCompany FundamentalsCorporate EarningsAnalyst EstimatesInvestor Sentiment & PositioningTrade Policy & Supply Chain
This Artificial Intelligence (AI) Stock, Up 28,700% Since Its IPO, Could Be the Biggest Bargain of the Decade

Surging demand from hyperscalers for AI memory and storage is accelerating Micron’s revenue and margins as the company benefits from high-bandwidth memory (HBM) adoption; the HBM TAM is forecast to grow at a ~40% CAGR to $100 billion by 2028. Micron generated roughly $10 EPS last year, with Street consensus at $33.20 for the current year and a forward P/E near 12; the author argues that re-rating to a doubled forward P/E could imply a ~100% upside to $780/share, positioning Micron as a deeply discounted play on AI infrastructure.

Analysis

Market structure: Hyperscalers (AMZN, GOOGL, MSFT) and HBM/storage suppliers (Micron MU, plus ASML/lam-retooling suppliers downstream) are primary winners as capex shifts from GPUs alone to memory-intensive stacks; GPU vendors (NVDA, AMD) still win but face less incremental share of marginal AI spend. Pricing power for HBM/advanced DRAM is improving because capacity is supply-inelastic near-term (12–24 month lead times) and TAM for HBM is modeled ~40% CAGR to $100B by 2028, implying multi-year above-trend ASPs if demand holds. Risk assessment: Tail risks include accelerated US export controls on advanced memory to China, a hyperscaler inventory correction (sudden 30–60% order drop), or rapid capex pullback that could halve DRAM pricing within 6–12 months. Watch immediate catalysts (next 1–3 months): NVDA/hyperscaler capex commentary and MU earnings/inventory cadence; medium-term (3–12 months) depends on ASP trends and wafer fab capacity additions. Trade implications: Base case — MU is underpriced (forward P/E ~12 vs peers); actionable trades are medium-term long MU via LEAPs or call spreads while hedging near-term earnings risk; consider relative-value longs in MU vs shorts in legacy NAND/consumer-cycle names (Samsung SSNLF / SK Hynix 000660.KS) to isolate HBM exposure. Cross-asset: stronger AI capex can tighten credit spreads for capex-financing corporates and lift equipment vendors; implied vols will spike around earnings/guide events — use time-limited options to control gamma. Contrarian angles: Consensus underestimates industry cyclicality and customer concentration — valuation catch-up assumes no policy shock or vertical integration by hyperscalers. Historical DRAM cycles show +100% rallies reversing >50% in 6–12 months after overbuild; therefore size positions conservatively, require confirmation of two consecutive quarters of ASP stability before increasing exposure aggressively.