$1,000 Invested in Micron Before Sept. 30 Could Be Worth This Much by 2028

Micron is positioned to benefit from an AI-driven memory-chip supply shortage that management expects could persist until 2028, with new capacity not scheduled to come online until mid-2027 and 2028. The article argues that even if memory pricing peaks next year and margins normalize to 30%-40%, sustained AI infrastructure spending could support strong earnings; it cites an estimate of $248 billion in next-year revenue and a potential $1.98 trillion valuation at a 40% margin and 20x earnings multiple. The bullish thesis hinges on Micron's Sept. 30 FY2026 Q4 results and any indication that supply tightness will last longer than investors expect.

Analysis

The relevant earnings question is not whether MU prints another beat, but whether its HBM allocation, DRAM contract-price trajectory, and customer concentration support estimates after the next two quarters. A sustained tight market shifts memory from a spot-price commodity narrative toward a capacity-allocation narrative, supporting a higher through-cycle multiple; however, MU remains exposed to faster-than-expected qualification gains at Samsung Electronics and SK hynix, which would weaken its pricing power before greenfield capacity arrives. NAND is the weak link: it is less AI-specific and a recovery there should not be capitalized at the same multiple as HBM/leading-edge DRAM.

The article's revenue, market-cap, and margin assumptions are not internally credible for MU and should not inform a valuation decision. The actionable verification set before earnings is: HBM bit-growth and yield commentary, DRAM/NAND pricing guidance, gross-margin bridge, capex intensity, and whether management commits incremental supply beyond previously disclosed plans. A beat driven by one-time pricing or mix, without upward FY27 supply-demand guidance, is likely to fade after an initial move because consensus is already positioned for elevated profitability.

Near term, MU earnings can also read through to NVDA: confirmed memory availability removes a potential bottleneck for accelerator-system shipments and supports server OEM build plans. Conversely, an aggressive MU capacity response would be more damaging to memory equities than to NVDA; lower memory costs ultimately improve AI-system economics for hyperscalers and platform suppliers. Over 6-18 months, the cleaner relative expression is quality HBM exposure versus commodity NAND exposure rather than a blanket long-memory trade.

Contrarian risk is that investors are extrapolating industry tightness while underweighting the typical 12-24 month lag between price signals and capacity additions. The thesis is falsified by sequential DRAM contract-price deceleration, HBM inventory normalization at hyperscalers, a material capex increase from MU/Samsung/SK hynix, or gross-margin guidance that fails to rise despite higher revenue.

Key Decisions for Investors

• Treat MU as an earnings-event watch rather than chase exposure pre-report: initiate only if guidance raises the next two-quarter gross-margin or HBM-revenue outlook and the stock holds the post-earnings gap for 1-2 sessions; target a 10-15% upside over 1-3 months, with a 7-8% stop on evidence the beat was pricing-only.
• Express the supply-tightness thesis with a 3-6 month long MU / short WDC pair, sized beta-neutral. MU has superior AI-memory sensitivity, while WDC is more exposed to cyclical storage pricing; exit if NAND pricing accelerates relative to DRAM or MU signals materially higher commodity-capacity additions.
• For NVDA, use MU's HBM commentary as a confirmation indicator rather than a standalone catalyst: add only if MU identifies improving supply availability without pricing-driven demand destruction. Reduce NVDA exposure if memory availability is cited as constraining system shipments or hyperscaler deployments.
• Set alerts for sequential DRAM contract-price growth turning negative or a combined capex escalation from the major memory producers. Either signal would warrant reducing MU within days, since multiple compression generally precedes the eventual earnings impact by several quarters.