Micron Technology — FQ3 FY2026 Earnings Preview

FQ3 FY2026 = Quarter ended ~May 29, 2026  ·  Reports Tuesday, July 1, 2026 AMC  ·  Prepared May 26, 2026  ·  FYE late August  ·  36 days out  ·  Stock $902.88

Micron reports FQ3 FY2026 on Tuesday, July 1 AMC. This is the most consequential memory semiconductor print in a decade. FQ2 was a blowout of historic proportions: revenue of $23.9B (+196% YoY, +75% QoQ) crushed the $18.7B guide by $5.2B, gross margin hit 75% (vs 68% guide, +700bps above midpoint), and EPS of $12.20 beat the $8.42 guide by 45%. Management then guided FQ3 to $33.5B revenue — a single quarter that exceeds Micron's entire annual revenue for every year through FY2024. The 81% gross margin guide is unprecedented for any memory company in history. DRAM supply remains severely constrained (50-67% fulfillment rate, now in its third consecutive quarter), NAND pricing is accelerating faster than DRAM, and the first five-year Strategic Customer Agreement (SCA) signals a structural shift from cyclical commodity to durable AI infrastructure franchise. The stock hit $1T market cap today on a UBS triple of their price target to $1,625.

Bull case: The $33.5B guide is mathematically conservative based on the accelerating beat pattern: FQ4 2025 revenue beat by +$0.6B, FQ1 2026 by +$1.1B, FQ2 2026 by +$5.2B. Even applying FQ1-level conservatism implies $35B+; applying FQ2-level implies $39B+. DRAM contract prices are up 58-63% QoQ in Q2 CY2026, and NAND prices are up 70-75% QoQ — both well above what the guide assumes. The 81% gross margin guide could prove conservative given FQ2's 700bps beat above midpoint. HBM4 volume shipments for NVIDIA Vera Rubin are underway with yields maturing faster than HBM3E. The first five-year SCA structurally de-risks the earnings stream — UBS projects EPS exceeding $100 through 2029. FCF could "roughly double sequentially" per CFO Murphy, implying $13B+ in a single quarter. CHIPS Act buyback restrictions expire in December 2026, unlocking potentially $50B+ in cash for shareholder returns. No new industry greenfield capacity arrives before late CY2027 at the earliest.

Bear case: Stock at $903 = ~$1T market cap at ~8x forward earnings — but that forward P/E is deceptive because it assumes peak-cycle earnings are sustainable. An 81% gross margin exceeds even NVIDIA's ~75%, and history says memory margins mean-revert aggressively. The FQ2 print triggered a 30% sell-off over six sessions despite the blowout beat, driven by peak margin fears and a Google TurboQuant paper showing 6x KV-cache memory compression. FY2027 CapEx steps up by $10B+ in construction spend alone — investment risk if the cycle turns. PC and smartphone units are declining low-double-digits due to price-driven demand destruction. Non-HBM margins are currently higher than HBM margins (a surprise that complicates the bull narrative as HBM mix grows). Samsung's 18-day strike creates near-term volatility. Harvard's Willy Shih warns "this too will pass" — new capacity from Samsung P5, SK Hynix M15X, and Micron's own Idaho/Tongluo fabs arrives in late 2027/2028. At $1T, any deceleration in the beat magnitude could trigger a sharp re-rating.

What's at stake: Confirmation that (a) revenue prints above $33.5B guide (consensus $34.9B), (b) gross margin holds at or above 81%, (c) FQ4 guide sustains the trajectory toward $37-40B (implied by FY2026 consensus of ~$108.7B), (d) additional SCA announcements beyond the first five-year deal, (e) HBM4 yield and mix color, (f) any buyback authorization preview ahead of CHIPS restriction expiry, (g) NAND pricing sustainability — the forgotten bull catalyst. A clean beat-and-raise with SCA progress likely takes the stock toward $1,000+; a soft guide or any hint of demand moderation could re-test $700.

Guidance issued on FQ2 2026 earnings call, March 18, 2026. MU guides revenue, gross margin, OpEx, and EPS for Q+1. Consensus estimates from ChartMill, MarketBeat, Simply Wall St.

Revenue has nearly tripled in three quarters ($11.3B to $23.9B to guided $33.5B). Gross margin expanded from 46% to 57% to 75%, now guided to 81%. EPS has gone from $3.03 to $12.20, guided to $19.15. FCF went from $0.8B to $6.9B and is projected to "roughly double" in FQ3 (~$13B+). Operating leverage is extraordinary: OpEx grew only 17% over three quarters while revenue tripled. DRAM remains ~79% of revenue; NAND is the under-appreciated accelerator with prices up 70-75% QoQ. Cells link to Daloopa source pages.

Each row shows guidance issued at the earnings call for the next quarter, matched against the actual result when that quarter was reported.

Key takeaway: Micron has beaten the high end of its own guidance range on revenue, EPS, and gross margin in every quarter for three consecutive quarters — and the beat magnitude is accelerating dramatically. Revenue beats expanded from +$0.6B to +$1.1B to +$5.2B. EPS beats from +21% to +28% to +45%. GM beats from +370bps to +530bps to +700bps. The guidance framework has consistently lagged the pace of DRAM/NAND pricing improvement. Even applying FQ1-level conservatism (the more moderate quarter) to the FQ3 guide implies revenue of $35B+ and EPS of $24+. The consensus at ~$34.9B/$19.88 reflects expectation of a beat but likely still underestimates the magnitude given the pattern.

Post-Guidance Updates (Since March 18, 2026):
March 15: Completed Tongluo P5 fab acquisition in Taiwan ($1.8B) — ahead of schedule, ~300K sq ft of 300mm cleanroom.
May 20: JPM Conference presentation — reiterated shortage duration and $200B US investment plan.
May 22: Mehrotra at TiEcon: "We see this shortage continuing beyond, well beyond 2026 timeframe... meaningful new supply in the industry doesn't really start ramping until 2028."
May 22: Began 1-alpha DRAM production at Manassas, Virginia — "most advanced memory ever made in the U.S." — $2B+ expansion.
May 26: UBS triples PT to $1,625 (from $535); stock surges 19% to $903, crossing $1T market cap.
Ongoing: Samsung 18-day strike by ~45,000 workers — TrendForce estimates 3-4% hit to global DRAM supply, accelerating order shifts toward Micron.

Key Q&A Themes from FQ2 Call:
1. SCA Terms & Downside Protection — Every analyst pressed on this. Sanjay refused specifics but emphasized "robust provisions" and "specific commitments." The first SCA is five years and spans "multiple markets." The bull case requires these to function as margin floors; the bear case is that they are rebranded LTAs.
2. Gross Margin Sustainability at 81% — Krish Sankar, Vivek Arya both pushed. Murphy's response: historical mean-reversion framework should be "revisited" because AI is a "transformational secular driver." Not guiding higher but not signaling peak either.
3. HBM Competitive Position — Sanjay confirmed share reached DRAM parity in CY25. Crucially, non-HBM margins are currently HIGHER than HBM margins — the opposite of consensus assumption. Refuses to break out HBM share going forward.
4. Capital Allocation — Tim Arcuri asked about $35-40B projected FCF and what to do with $50B+ in cash. Return on capital heading "towards 50%." Buyback capacity massive but CHIPS-restricted until Dec 2026.
5. Consumer Demand Destruction — PC/smartphone units declining "low double-digit percentage range" in CY26. Management says it is "accounted for in our forecast" and offset by content growth (flagships at 12GB+ DRAM, AI PCs at 32GB+).
6. New Architectures (LPU/SRAM) — Asked about NVIDIA Grok LPX displacing HBM. Sanjay reframed as complementary: LPU uses 12TB DDR5 per rack and works alongside Vera Rubin. "Any architecture that makes AI infrastructure more efficient is good for AI."

Highest-conviction catalyst: Additional SCA announcements at or before FQ3 earnings. This is the single factor that could change how the market models Micron's earnings durability and drive further multiple expansion from the current ~8x forward P/E toward 12-15x.

Key takeaway: MU is a serial earnings beater with an accelerating beat trajectory. EPS beats have expanded from +2-10% range (FQ3 2023 - FQ1 2025) to +20-39% range (FQ3 2025 - FQ2 2026). Revenue beats have followed the same pattern, from +0-2% to +5-19%. The operating leverage effect is the most under-modeled dynamic: when revenue beats, the EPS flow-through is massively larger than the street expects because fixed costs are a shrinking share of revenue at these scales. The base case for FQ3 2026 is another beat on both EPS and revenue. The key risk to the streak would be a demand air pocket or inventory correction — neither is currently signaled.

MU reports first among all listed peers. No direct peer results available for read-through before the July 1 print.

1. FQ3 Revenue vs. $33.5B guide: Consensus $34.9B. The accelerating beat pattern (last 3Q: +6%, +9%, +28% above guide mid) implies $36B+ is plausible. Below $33.5B would break the streak and trigger a sharp sell-off.

2. Gross margin vs. 81% guide: The most-watched metric. Last 3Q beat guide by +370bps, +530bps, +700bps. But at 81% the law of diminishing returns applies — even bulls expect smaller beats. 83%+ would be extraordinary; below 81% would concern.

3. FQ4 guidance — the next leg: FY2026 consensus of ~$108.7B implies FQ4 needs to be $37-40B. The guide will tell us if the sequential ramp sustains. Any guide below $35B would disappoint.

4. SCA progress: How many additional SCAs signed beyond the first five-year deal? Any disclosure on pricing mechanisms or margin floors? This is the single most important structural catalyst for re-rating.

5. HBM4 yield and mix: What is the HBM4 vs. HBM3E revenue split in FQ3? Is HBM4 margin-accretive yet? Any update on HBM4E development timeline?

6. NAND pricing sustainability: Can 70%+ QoQ price increases persist? NAND has been the under-appreciated margin story — any commentary on data center SSD momentum is key.

7. Buyback authorization: CHIPS restrictions expire Dec 2026. Will the board pre-announce a buyback program? With potentially $50B+ cash, this is a significant capital return catalyst.

8. Free cash flow magnitude: Murphy projected FCF could "roughly double sequentially" from FQ2's $6.9B — implying $13B+ in a single quarter at ~$7B CapEx.

9. FY2027 CapEx magnitude: Construction spend >$10B higher YoY. Total FY27 CapEx could be $35B+. Will the construction-heavy mix weigh on returns?

10. Supply shortage duration: Does Mehrotra maintain the "beyond 2026" language? The 50-67% fulfillment rate is in its 3rd consecutive quarter — any softening would be a major signal.

11. Tariff update: Guidance explicitly excludes tariff impact. Any new color on surcharge pass-through or demand disruption from trade policy?

12. Samsung strike impact: 18-day walkout could shift orders to MU. Any commentary on incremental demand pull-forward or customer diversification away from Samsung?

Valuation Snapshot: At $902.88, MU trades at ~8.0x forward P/E (FY2027E ~$98 EPS) and ~35x trailing P/E ($21.26 TTM EPS). The forward multiple is deceptively low — it assumes peak-cycle earnings are sustainable. For context: SK Hynix (000660.KS) trades at ~7-8x forward with 72% operating margins in Q1 2026 and ~57-62% HBM share vs MU's ~21%. Samsung Memory trades at a conglomerate discount. The key valuation debate: should MU trade at a traditional memory cyclical multiple (7-10x) or re-rate toward AI infrastructure multiples (15-20x)? UBS's $1,625 PT implies ~15x CY2029E EPS of $117 discounted one year — a structural re-rating thesis predicated on SCAs making earnings durable.

FQ3 Segment Estimates (Directional): Based on FQ2 mix (DRAM 79%, NAND 21%) and guided revenue of $33.5B midpoint: DRAM ~$26.5B (driven by HBM4 ramp + continued ASP increases), NAND ~$7.0B (DC SSD growth continuing). By business unit: CMBU (Cloud Memory) ~$10-11B, CDBU (Core Data Center) ~$8-9B, MCBU (Mobile & Client) ~$10-11B, AEBU (Auto & Embedded) ~$3.5-4B. CDBU likely shows the strongest QoQ growth given DC SSD momentum (+139% QoQ in FQ2).

Management Quality: CEO Sanjay Mehrotra (since May 2017, 9 years) has overseen Micron's transformation from a commodity memory maker to an AI infrastructure franchise. CFO Mark Murphy (since 2019, 7 years) has been a steady hand on capital allocation, delivering credit upgrades to BBB and managing the balance sheet from net debt to $6.5B net cash. Guidance accuracy is exceptional — 12/12 EPS beats with accelerating magnitude over the last 3 years. C-suite stability is strong with no senior departures. Management's track record supports the "conservative guide, beat-and-raise" pattern that investors can underwrite.