Micron Technology
Rating
Accumulate
Adding on Dips — Active Accumulation
Combined average of Moat (AI Resilience), Growth, and Valuation scores.
Moat Score
An oligopoly of three (Samsung, SK Hynix, Micron) with high capital barriers to entry, but commodity memory pricing has historically limited moat durability. That constraint is now being rewritten: sixteen Strategic Customer Agreements (SCAs) — take-or-pay, typically five years through 2030, with ~$100B of disclosed RPO at minimum contract pricing and $22B of projected customer cash deposits — convert a meaningful share of Micron's book from quarterly spot negotiation into contracted volume and floor pricing. HBM4 is in high-volume shipment for the lead customer platform, and non-GAAP gross margin printed 84.9% in Q3 FY2026 with Q4 guided to ~86%.
Micron's competitive position rests on Oligopoly Structure, HBM4 Execution, and SCA Lock-In — the last of these crossed from narrative to contracted fact in the June 24 print:
- Three-Player Oligopoly: With Samsung, SK Hynix, and Micron controlling ~95% of DRAM supply, the market is structurally oligopolistic. New entrants face $30B+ capex requirements and decade-long learning curves that effectively preclude competition. Micron is the only US-based survivor of what was once a much larger industry.
- HBM4 Shipping — Margin Proof at Scale: HBM4 on 1β DRAM is in high-volume shipment for the lead customer's platform, with qualification samples at multiple end-customers and HBM4E (1γ) volume production expected in calendar 2027. Q3 FY2026 revenue of $41.46B (+346% YoY) and non-GAAP gross margin of 84.9% — with Q4 guided to ~$50B and ~86% — confirm that the HBM mix is delivering peak-cycle economics at scale, not just design-win headlines.
- Strategic Customer Agreements: Sixteen SCAs spanning data center, consumer, and automotive — roughly 20% of DRAM volume and a third of NAND over the term — carry take-or-pay volume commitments, price floors/ceilings, and non-cancellation terms. Fourteen of them alone disclose ~$100B of cumulative minimum-price RPO, with $22B of projected customer financial commitments (~$18B cash deposits). Management targets half or more of company revenue under SCAs when the program is complete. Outside the contracted book, standard DRAM and NAND remain cyclical, and the elevated capex run-rate still creates overcapacity risk if AI demand normalises beyond the SCA floor.
Ten Moats Verdict
Micron is a clear net beneficiary of AI — the HBM4 supercycle is directly driven by AI infrastructure build-out, talentScarcity and proprietaryData are strengthened by AI's demand for specialised chip design, and the SCA program has upgraded transactionEmbedding from intact to strong with ~$100B of contracted RPO and $22B of customer deposits. The moat is materially better than a year ago and better than the July 10 file captured. Durability still hinges on SCA coverage expanding toward the ≥50%-of-revenue target and on the HBM margin premium holding through CY2027–2028; Micron does not own a software layer, a data flywheel, or a network effect that compounds independently of the hardware cycle.
Micron is a B2B semiconductor manufacturer with no consumer interface lock-in.
memory chips have no embedded business-logic moat.
Micron does not derive competitive advantage from public data access.
Leading-edge DRAM and HBM process engineers (sub-1β/1γ node specialists, HBM4 base-die architects, advanced metallization specialists) are among the scarcest technical talent globally. Micron's Boise R&D center is a decade-deep talent cluster that competitors cannot quickly replicate. AI strengthens this moat — designing HBM4/HBM4E base logic dies in-house requires irreplaceable human expertise.
Micron sells DRAM, NAND, and HBM as distinct products with limited bundling; some system-level memory solutions exist but don't create meaningful lock-in vs. Samsung or SK Hynix. SCAs span the portfolio for some customers, but that is contractual embedding rather than product-suite bundling.
Proprietary DRAM cell designs (1β/1γ nodes), HBM4 base-die CMOS architecture, advanced metallization processes, and yield-learning data from high-volume HBM production represent genuine IP. In-house logic die design (vs. competitors outsourcing) is a defensible advantage AI cannot easily replicate.
CHIPS Act $6.4B in total grants for Idaho and New York fabs makes Micron a designated US national security asset. The US government has an explicit interest in Micron's success as the only US-based DRAM manufacturer — and export controls on Samsung/SK Hynix to China further entrench Micron's strategic position.
no network effects exist in commodity memory; customers buy on price, availability, and quality specifications, not ecosystem lock-in.
Upgraded intact → strong on the June 24 SCA disclosure. Sixteen take-or-pay Strategic Customer Agreements — typically five years through 2030, non-cancellable, with volume commitments and price floors/ceilings — disclose ~$100B of cumulative minimum-price RPO and $22B of projected customer financial commitments (~$18B cash deposits). That is concrete, named switching cost: walking away means forfeiting deposits and still paying for committed volume. Coverage is still partial (~20% of DRAM, ~⅓ of NAND; management targets ≥50% of revenue), so this is deepening rather than a completed fortress, but it is no longer a soft 'multi-year LOI' claim.
memory is a commodity input; Micron is not a system of record for any business function; customers source from all three suppliers simultaneously.
Combined average of Moat (AI Resilience), Growth, and Valuation scores.
Moat Score
An oligopoly of three (Samsung, SK Hynix, Micron) with high capital barriers to entry, but commodity memory pricing has historically limited moat durability. That constraint is now being rewritten: sixteen Strategic Customer Agreements (SCAs) — take-or-pay, typically five years through 2030, with ~$100B of disclosed RPO at minimum contract pricing and $22B of projected customer cash deposits — convert a meaningful share of Micron's book from quarterly spot negotiation into contracted volume and floor pricing. HBM4 is in high-volume shipment for the lead customer platform, and non-GAAP gross margin printed 84.9% in Q3 FY2026 with Q4 guided to ~86%.
Growth Score
Micron is in the steepest part of a semiconductor supercycle driven by AI memory demand. Q3 FY2026 (ended May 28, reported June 24) delivered record revenue of $41.46B (+346% YoY, +74% sequential) and non-GAAP EPS of $25.11 — crushing the prior ~$33.5B / ~$19.15 guide and ~$35.8B / ~$20.7 Street. Non-GAAP gross margin reached 84.9%. Management guided Q4 FY2026 to ~$50.0B revenue, ~86% gross margin, and ~$31.00 non-GAAP EPS, which puts the FY2026 non-GAAP EPS trajectory at roughly $73 (Q1 $4.78 + Q2 $12.20 + Q3 $25.11 + Q4 ~$31) on ~$129B of revenue. The structural disclosure was the SCA program: 16 take-or-pay multi-year agreements with ~$100B RPO and $22B of customer commitments, and HBM TAM now expected to cross $100B in fiscal 2027 (pulled forward from 2028). Since the prior July 10 refresh — which still carried stale Q2 metrics and the obsolete $33.5B Q3 guide — the stock ran above $1,200 post-print, sold off ~29% in July on profit-taking and China memory-competition fears, and has recovered to ~$893. SK Hynix's Nasdaq listing and the June DRAM price-fixing class action remain secondary overlays on a thesis now dominated by contracted AI-memory demand.
Valuation Score
At ~$893 (market cap ~$1.01T), MU has given back the post-Q3 spike above $1,200 — including a ~29% July drawdown — and now sits just below the revised $950 base case. On the FY2026 non-GAAP EPS trajectory of ~$73, forward P/E is ~12×; on TTM non-GAAP EPS of ~$45 the trailing multiple is ~20×. That is a material re-rating of the earnings power versus the July 10 file (which still priced ~$58 FY26 EPS at ~$1,021 / ~18×), and the SCA disclosure is why the scenario ladder moved up with it. PEG ~0.4× on a 25–35% blended CAGR still signals growth at a reasonable price, but memory-cycle history and the incomplete SCA coverage (~20% of DRAM) argue against treating ~$73 of peak-cycle EPS as the permanent base.
Oligopoly with Contracted Walls
Micron's competitive position rests on Oligopoly Structure, HBM4 Execution, and SCA Lock-In — the last of these crossed from narrative to contracted fact in the June 24 print:
- Three-Player Oligopoly: With Samsung, SK Hynix, and Micron controlling ~95% of DRAM supply, the market is structurally oligopolistic. New entrants face $30B+ capex requirements and decade-long learning curves that effectively preclude competition. Micron is the only US-based survivor of what was once a much larger industry.
- HBM4 Shipping — Margin Proof at Scale: HBM4 on 1β DRAM is in high-volume shipment for the lead customer's platform, with qualification samples at multiple end-customers and HBM4E (1γ) volume production expected in calendar 2027. Q3 FY2026 revenue of $41.46B (+346% YoY) and non-GAAP gross margin of 84.9% — with Q4 guided to ~$50B and ~86% — confirm that the HBM mix is delivering peak-cycle economics at scale, not just design-win headlines.
- Strategic Customer Agreements: Sixteen SCAs spanning data center, consumer, and automotive — roughly 20% of DRAM volume and a third of NAND over the term — carry take-or-pay volume commitments, price floors/ceilings, and non-cancellation terms. Fourteen of them alone disclose ~$100B of cumulative minimum-price RPO, with $22B of projected customer financial commitments (~$18B cash deposits). Management targets half or more of company revenue under SCAs when the program is complete. Outside the contracted book, standard DRAM and NAND remain cyclical, and the elevated capex run-rate still creates overcapacity risk if AI demand normalises beyond the SCA floor.
Ten Moats Verdict
Micron is a clear net beneficiary of AI — the HBM4 supercycle is directly driven by AI infrastructure build-out, talentScarcity and proprietaryData are strengthened by AI's demand for specialised chip design, and the SCA program has upgraded transactionEmbedding from intact to strong with ~$100B of contracted RPO and $22B of customer deposits. The moat is materially better than a year ago and better than the July 10 file captured. Durability still hinges on SCA coverage expanding toward the ≥50%-of-revenue target and on the HBM margin premium holding through CY2027–2028; Micron does not own a software layer, a data flywheel, or a network effect that compounds independently of the hardware cycle.
Micron is a B2B semiconductor manufacturer with no consumer interface lock-in.
memory chips have no embedded business-logic moat.
Micron does not derive competitive advantage from public data access.
Leading-edge DRAM and HBM process engineers (sub-1β/1γ node specialists, HBM4 base-die architects, advanced metallization specialists) are among the scarcest technical talent globally. Micron's Boise R&D center is a decade-deep talent cluster that competitors cannot quickly replicate. AI strengthens this moat — designing HBM4/HBM4E base logic dies in-house requires irreplaceable human expertise.
Micron sells DRAM, NAND, and HBM as distinct products with limited bundling; some system-level memory solutions exist but don't create meaningful lock-in vs. Samsung or SK Hynix. SCAs span the portfolio for some customers, but that is contractual embedding rather than product-suite bundling.
Proprietary DRAM cell designs (1β/1γ nodes), HBM4 base-die CMOS architecture, advanced metallization processes, and yield-learning data from high-volume HBM production represent genuine IP. In-house logic die design (vs. competitors outsourcing) is a defensible advantage AI cannot easily replicate.
CHIPS Act $6.4B in total grants for Idaho and New York fabs makes Micron a designated US national security asset. The US government has an explicit interest in Micron's success as the only US-based DRAM manufacturer — and export controls on Samsung/SK Hynix to China further entrench Micron's strategic position.
no network effects exist in commodity memory; customers buy on price, availability, and quality specifications, not ecosystem lock-in.
Upgraded intact → strong on the June 24 SCA disclosure. Sixteen take-or-pay Strategic Customer Agreements — typically five years through 2030, non-cancellable, with volume commitments and price floors/ceilings — disclose ~$100B of cumulative minimum-price RPO and $22B of projected customer financial commitments (~$18B cash deposits). That is concrete, named switching cost: walking away means forfeiting deposits and still paying for committed volume. Coverage is still partial (~20% of DRAM, ~⅓ of NAND; management targets ≥50% of revenue), so this is deepening rather than a completed fortress, but it is no longer a soft 'multi-year LOI' claim.
memory is a commodity input; Micron is not a system of record for any business function; customers source from all three suppliers simultaneously.
Growth Analysis
Growth Drivers
Key Risk
If AI hyperscaler capex enters a pause cycle in 2H CY2027–FY2028, Micron's elevated capex run-rate still creates overcapacity risk outside the SCA book. SCAs cushion the floor — take-or-pay volume at price bands management says keep gross margin well above prior-cycle peaks — but they cover only ~20% of DRAM and ~⅓ of NAND so far, and renegotiation pressure in a deep downcycle is untested. Falsifiable: FY2028 non-GAAP EPS reverting below $35, or gross margin compressing below 50% for two consecutive quarters, would break the 'contracts rewrite cyclicality' claim. Secondary: the June 2026 DRAM price-fixing class action (Samsung, SK Hynix, Micron) remains early-stage with no class certified.
Score Derivation
90.0 base + 2.7 trajectory + 4 margin − 5 risk = 92
Base 90 (25–35% blended CAGR midpoint, decayed from the +346% Q3 print toward SCA-supported mid-cycle) + ~3 trajectory (2 of 3 drivers accelerating) + 4 margin expansion (non-GAAP GM 84.9% → ~86% guided) − 5 moderate cyclical/legal risk ≈ 92
HBM + SCA Demand Structural Shift
Price Scenarios (12–24 Months)
Valuation Analysis
At ~$893, Micron trades at ~12× the ~$73 FY2026 non-GAAP EPS trajectory and ~20× TTM — inside the historical memory-peak band (8–18× forward) and well below software peers. The SCA floor (price bands that management says protect gross margin above prior-cycle peaks, plus $22B of customer commitments) supports paying a higher trough multiple than 2022's wipeout implied, which is why the bear case moved to $450 rather than a sub-$300 revisit. With spot roughly at the $950 base, this is a fair-value hold on mid-cycle assumptions rather than a deep-value entry — add on weakness toward the mid-$700s, not chase a re-test of $1,200 without another guide raise. ~$950.
Valuation Multiples
| Trailing P/E (non-GAAP) | ~20× |
| Forward P/E (FY26) | ~12× |
| PEG Ratio | ~0.4× |
| Price / Sales (FY26) | ~7.8× |
| Price / FCF | ~14× |
Forward P/E has compressed to ~12× as the June 24 beat-and-raise re-rated FY2026 EPS from the old ~$58 path to ~$73, while the stock itself retreated from ~$1,021 / ~$1,255 peaks to ~$893. PEG ~0.4 still signals growth at a reasonable rate. The risk is unchanged in kind: if HBM oversupply emerges into FY28 outside the SCA floor, EPS could revert toward $30–35 and a 13× trough multiple implies $390–455 — which is why the bear case sits there rather than at the old cycle-trough multiples on unprotected spot pricing.
Approximate figures as of August 2026.
Where We Are vs Targets
Loading live price…
Memory cycle reversion: AI hyperscaler capex pauses in 2H CY2027, Samsung and SK Hynix flood HBM capacity, and SCAs cushion but do not prevent a sharp downcycle outside the contracted book.
- AI hyperscaler capex pause in 2H CY2027 reduces incremental HBM demand; non-SCA pricing collapses
- Samsung and SK Hynix close the HBM4 yield gap; Micron's CY2027–28 share advantage erodes
- Standard DRAM and NAND prices fall 30%+ outside take-or-pay floors; FY28 EPS reverts toward $30–35
- Multiple compresses to ~13× trough earnings → ~$450 implies ~$35 EPS at 13×, with SCAs preventing a 2022-style sub-$20 EPS wipeout
HBM supercycle and SCA coverage sustain through CY2027; Micron delivers FY2026 EPS of ~$73, and contracted floors hold mid-cycle gross margins well above prior peaks into FY2027–28.
- FY2026 revenue tracks toward ~$129B as the ~$50B Q4 guide materialises
- SCA program expands toward management's ≥50%-of-revenue target; HBM maintains competitive share through CY2027
- FY2027–28 EPS settles in a $55–70 mid-cycle band as HBM trade ratios and SCA floors limit the amplitude of the next down-leg
- Stock trades at ~14–15× mid-cycle EPS (~$950 on ~$65) — a premium to unprotected memory peaks justified by contracted visibility
SCAs reach ≥50% of revenue, HBM becomes the dominant AI inference memory, and gross margins permanently re-rate above 70% even off-peak.
- HBM4E/HBM5 commitments and completed SCA program entrench Micron with NVIDIA and the hyperscalers through 2030
- AI inference at scale creates a second demand wave — HBM TAM exceeds the $100B FY2027 mark and keeps growing
- Near-memory compute integration (PIM/CXL) opens adjacent TAM, further differentiating Micron from commodity DRAM
- FY2027 EPS reaches $90–100; stock re-rates to ~14–15× as contracted HBM is treated less like memory, more like accelerator content