Back to News
Market Impact: 0.25

Can Micron Technology Become the New Nvidia?

Technology & InnovationCredit & Bond MarketsCorporate EarningsAnalyst EstimatesMarket Technicals & FlowsCompany FundamentalsCapital Returns (Dividends / Buybacks)
Can Micron Technology Become the New Nvidia?

The article highlights that Micron (MU) is riding sustained DRAM/NAND tight supply into 2027+, with DRAM at ~76% of sales and demand supported by AI data-center buildouts (hyperscalers targeting ~$650B in capex this year, potentially >$1T next year). Micron is cited as having posted another “jaw-dropping” quarter and expecting next-quarter revenue of about $50B, implying strong near-term momentum despite the AI-stock selloff. Valuation scenarios using FY2027 EPS of $149.64 (range $149.64–$221.27) suggest a potential $2,244–$3,741 share price at 15x–25x earnings (MC ~$2.5T–$4.2T), with a high-end case up to ~$6.2T if earnings and multiples both run.

Analysis

Micron is the cleaner second derivative on AI capex than Nvidia: every incremental server rack increases memory attach rates, but MU has much higher operating leverage because ASPs, not unit growth, drive earnings inflection. The key market mechanism is that memory supply response is lagged by quarters to years, so if hyperscaler spending stays elevated through 2026, MU can keep compounding even if the broader AI basket de-rates on multiple compression.

The risk is that this is also the most cyclical way to play AI. If hyperscaler capex pauses for even one budget cycle, memory prices can normalize before new wafer starts arrive, and MU’s earnings power will reset far faster than consensus models that extrapolate peak margins into FY27. The main second-order loser is not just slower cloud buyers; it is every downstream hardware assembler and OEM with inventory exposure, because memory cost inflation eventually forces either margin compression or bill-of-materials redesign.

Contrarian view: the market is treating the shortage as a quasi-structural scarcity, but the relevant question is not whether demand is strong today — it is whether capacity additions in 2027 create a visible overhang before investors have time to haircut estimates. If DRAM/NAND pricing weakens while FY27 consensus still moves higher, the multiple expansion thesis breaks. For now, the setup favors owning MU on pullbacks, but only with a hard watch on memory ASPs and hyperscaler capex revisions over the next 1-3 quarters.

More News