Micron Technology — Memory-Cycle Leverage Meets AI-Driven Demand

Micron Technology is a global manufacturer of memory and storage products, operating primarily in DRAM (dynamic random-access memory) and NAND flash. The company designs, manufactures, and sells these products through vertically integrated fabs and packaging facilities concentrated in the United States, Japan, Singapore, Taiwan, and other locations.

How it makes money: Micron sells memory components and modules to OEMs, hyperscalers, and distributors. Revenue is denominated in bits shipped multiplied by average selling price per bit. Because bit supply is fixed in the short run by fab capacity and wafer starts, ASP is the dominant swing factor in revenue and profitability.

Customer base: The customer set is concentrated among a relatively small number of large buyers — hyperscale cloud providers, server OEMs, smartphone manufacturers, PC OEMs, and automotive/industrial customers. Datacenter and AI-related demand has grown to represent a disproportionate share of the revenue mix and an even larger share of incremental profit.

Scale: With a market capitalization of $1,037.7B and 1,129.39M shares outstanding, Micron is one of the largest semiconductor companies globally. Its public float of 1,125.54M shares is nearly the entire share count, indicating minimal insider or strategic ownership overhang.

Near-term (next 4–8 quarters):

• HBM ramp. HBM capacity is sold out or heavily committed across the industry, and Micron's ability to convert wafer capacity to HBM determines near-term revenue growth and margin mix.
• Datacenter DDR5 and server refresh. General-purpose server memory demand follows hyperscaler capex, which remains elevated.
• Pricing momentum. Given the operating leverage in the model, ASP direction matters more than unit growth for near-term EPS.

Medium-term (2–5 years):

• AI inference at the edge. On-device AI in smartphones and PCs raises memory content per device, expanding the addressable market beyond datacenter.
• Automotive and industrial. These segments carry higher margins and longer design cycles, providing a partial hedge against consumer cyclicality.
• Capacity conversion economics. The ability to shift wafer starts between commodity DRAM, HBM, and specialty memory is a capital-allocation advantage that should smooth earnings over time.

Narrative: The table above is directional rather than a point forecast, because the single most important input — memory ASP — is not knowable with precision. What the table does illustrate is the shape of Micron's earnings distribution: wide, with meaningful dispersion around a mid-cycle midpoint. The trailing EPS of $44.28 sits above what we would characterize as mid-cycle, which is consistent with the current point in the cycle being favorable. Investors buying at $903.57 are paying roughly 20.4x trailing earnings; the relevant question is what multiple they are paying on normalized earnings, which we estimate is materially higher.

Market size / TAM: The global memory market (DRAM plus NAND) is a large, capital-intensive, oligopolistic industry measured in the hundreds of billions of dollars annually, with DRAM representing the larger and more profitable portion. AI infrastructure has expanded the effective TAM by increasing memory content per system and by creating the HBM category, which commands premium pricing.

Competitive positioning: Micron is one of three scaled DRAM manufacturers and one of a handful of NAND manufacturers. Its competitive position rests on process technology leadership, cost per bit, and — increasingly — HBM execution. Scale is essential: below a certain volume, fab economics do not work.

Named comparables:

• Samsung Electronics — the largest memory manufacturer by volume, with broad exposure across DRAM, NAND, and foundry.
• SK Hynix — the leading HBM supplier and Micron's most direct competitor in AI memory.
• Kioxia — a NAND-focused manufacturer, relevant for the storage side of the comparison.
• Western Digital — historically a NAND and HDD competitor, relevant for storage market dynamics.

DCF discussion: A discounted cash flow analysis for a memory manufacturer is unusually sensitive to the terminal-year assumption, because the business is cyclical and the terminal value dominates the output. We therefore anchor the DCF on normalized mid-cycle free cash flow rather than current-year cash flow, and apply a discount rate reflecting the beta of 2.22 — a high cost of equity that materially reduces the present value of distant cash flows. The result is a valuation range that is wide by design. Key sensitivities: (1) normalized gross margin, (2) HBM mix as a percentage of revenue, (3) capital expenditure intensity, and (4) the discount rate applied to a high-beta cyclical.

Comparable-company multiples:

The appropriate way to read this table is that Micron's multiple should be assessed against where it sits in the cycle, not against a static peer average. Peers at similar cycle points tend to trade at similar multiples; the dispersion appears when cycle positions diverge.

• Memory price cyclicality. ASP declines are the single largest driver of earnings deterioration. A sharp pricing correction would compress margins rapidly given the fixed cost base.
• HBM execution risk. HBM requires advanced packaging and stacking capabilities; yield issues or qualification delays at customers would delay revenue recognition and cede share to competitors.
• Customer concentration. A small number of hyperscalers and OEMs account for a large share of revenue; capex deferrals by any one of them would be material.
• Capital intensity. Leading-edge fabs require substantial, front-loaded investment. If demand weakens after capacity is committed, returns on that capital deteriorate sharply.
• Valuation and beta risk. With a beta of 2.22 and a 52-week range of $154.65–$1,255.00, MU is capable of large drawdowns. The recent -7.35% session on 8.33M shares illustrates this. Investors should size positions accordingly.