Will Micron Soar After June 24?

Micron posted record quarterly revenue of more than $23 billion, up nearly 200% year over year, with DRAM revenue up 207% and NAND revenue up 169%. Gross margin reached a company-record 75%, and the company forecast additional records for the fiscal third quarter. The article is constructive on Micron’s long-term AI exposure, but near-term upside may be tempered by supply constraints, a rich rally of more than 200% this year, and valuation at 16x forward earnings.

Analysis

MU is evolving from a cyclical memory name into a leveraged tollbooth on AI infrastructure intensity. The second-order dynamic is that every incremental layer of AI utility — training, inference, agentic workflows, retrieval, and persistent context — increases bytes moved and retained per compute cycle, which expands the TAM for high-bandwidth memory and server DRAM faster than unit GPU growth alone. That makes the debate less about whether AI capex slows and more about whether memory content per dollar of AI compute keeps rising; if it does, MU’s earnings power can remain elevated even if hyperscaler spending becomes lumpier.

The market’s main mistake is probably treating supply tightness as a near-term ceiling rather than a medium-term moat. In semis, scarcity usually compresses buyer optionality first and only later translates into pricing power destruction; here, constrained output may actually reinforce rational pricing discipline across the industry, which supports margins and keeps competitors from racing to the bottom. The flip side is that any sign of easing supply could trigger multiple compression faster than fundamentals deteriorate, because the stock is already discounting a prolonged shortage regime.

Near-term, the setup is event-driven rather than secular: the print can move the stock on guide quality, mix, and management commentary about 2025 capacity, but the bigger swing factor is whether investors infer acceleration in HBM adoption versus a plateau in conventional DRAM. A split announcement would matter mostly as a behavioral catalyst, not a fundamental one, and could attract momentum/retail flow into an already crowded winner. The real risk is a mild guidance miss paired with cautious capacity language, which would not break the AI thesis but could de-rate the stock 10-15% in a few sessions as profit-taking compounds.

Key Decisions for Investors

• Stay long MU into the event only via a defined-risk structure: buy MU calls or a call spread for the next 4-8 weeks, targeting a post-earnings upside pop from strong guide/possible split catalyst while capping downside if guidance is merely in-line.
• If already long MU spot, trim 20-30% into the print and re-add on any 8-12% post-earnings drawdown; the stock’s strongest risk/reward is likely on a post-event reset, not chasing a crowded pre-earnings extension.
• Pair trade: long MU / short INTC over 3-6 months. MU is directly monetizing AI memory content growth, while INTC is more exposed to slower-moving legacy server and PC cycles; this isolates the AI memory winner versus a structurally weaker beneficiary.
• For a cleaner AI infrastructure basket, consider long MU vs short NVDA on a relative-value basis for 1-2 quarters if AI capex becomes more selective. MU has less execution risk from customer concentration on product ramps, while NVDA is more vulnerable to any pause in hyperscaler spending multiples.
• Watch for evidence of 2025 capacity additions or HBM mix acceleration. If management signals supply normalization sooner than expected, exit longs quickly; the upside would still exist, but the multiple can compress before earnings power does.