Calls › MU › Aug 9, 2026

MU Lean bullish Micron Technology, Inc. · Technology

Memory demand appears sustainable over the next 3-6 months, particularly for HBM and server DRAM, but Micron's extraordinary revenue and margin growth is not eq

Published Aug 9, 2026 · Horizon 3-6 months

Confidence68/100

Entry (close)

$877.57 2026-08-07

Since publish

+5.8% to 2026-10-27

vs SPY

+6.3%

3-month verdict

Closes 2026-11-06

Thesis

Memory demand appears sustainable over the next 3-6 months, particularly for HBM and server DRAM, but Micron's extraordinary revenue and margin growth is not equally sustainable because it has been driven primarily by shortage pricing rather than comparable physical bit growth. The equity setup is modestly favorable: industry forecasts and customer commitments support another several quarters of tight supply, while the stock's roughly 8.6x multiple on the visible but potentially stale FY2027 EPS estimate already discounts substantial normalization. The asymmetry is positive but limited by extreme volatility, uncertain normalized earnings, and the possibility that decelerating prices are interpreted as a cycle peak.

Bull case

HBM and server DRAM are experiencing more than a conventional inventory rebound: AI infrastructure spending is still expanding, HBM consumes roughly three times the wafer capacity of DDR5, and major suppliers are reported sold out through 2027. Micron's long-term agreements, deposits, and minimum-volume commitments reduce near-term volume risk, while the shift of wafer capacity toward HBM constrains conventional DRAM supply. If Q4 guidance is achieved and FY2027 EPS remains near the visible estimate, the current multiple leaves room for appreciation even as price growth moderates.

Bear case

The headline growth rates exaggerate underlying consumption because Q3 DRAM revenue was driven by a low-60% sequential ASP increase against only low-single-digit bit growth. Contract-price increases are already forecast to decelerate, and memory equities often peak before reported earnings. Higher industry capex and fab ramps should eventually add supply; NAND may loosen before DRAM, while Bloomberg Intelligence and Bernstein commentary suggests shortage conditions peaked or begin easing during the second half of 2026. Strategic agreements can pull purchases forward, cap upside through ceiling prices, and create overbooking risk; they do not prove that end demand is incremental. If pricing rolls over before bit growth accelerates, Micron's very high margins and earnings estimates can fall quickly despite healthy AI demand.

What would change our mind

Micron misses the midpoint of fiscal Q4 revenue or EPS guidance for demand-related reasons rather than temporary execution issues.
Fiscal 2027 guidance shows declining DRAM bit shipments or a broad deterioration in data-center demand.
DRAM contract prices turn sequentially negative while lead times shorten and customer inventories rise.
HBM capacity previously described as sold out becomes available because customers defer or cancel deployments.
Hyperscalers materially cut aggregate AI infrastructure budgets rather than merely delay projects because of power constraints.
FY2027 EPS consensus falls materially below approximately $102 without a compensating decline in the share price; that would invalidate the low-multiple support for the equity thesis.
Micron weakens, renegotiates, or fails to collect expected deposits under the strategic customer agreements.

Since publication vs SPY · closed horizons are frozen

1 weekfinal

+10.3%

+10.7% raw · hit

1 monthfinal

+16.2%

+15.8% raw · hit

3 monthsrunning

+6.3%

+5.8% raw

6 monthsrunning

+6.3%

+5.8% raw

The full research everything the view rests on

Where the edge would come from

What the price assumes

At $877.57 on Aug. 7, 2026, MU had risen 129% in six months and 707% in one year but was 27.7% below its 52-week high. Using the visible approximately $102 FY2027 EPS estimate, the stock trades near 8.6x, suggesting the market expects near-term earnings to remain exceptional but then decline materially toward a lower mid-cycle level. In other words, the price does not assume current margins persist indefinitely, but it still assumes AI-related demand prevents a rapid memory downturn. Valuation evidence is imperfect because available consensus snapshots are inconsistent and some estimates predate the June earnings beat.

Why the market may be wrong

The market may be conflating slower memory-price increases with weakening end demand. TrendForce's forecast decelerates from Q2 price spikes to still-positive Q3 contract pricing, while independent forecasts indicate HBM and DRAM supply remains constrained through 2027. If Micron meets Q4 guidance and forward estimates remain stable while the shares remain roughly 15-30% below their high, the drawdown is more consistent with multiple compression than fundamental deterioration. The contrary possibility is that investors are overestimating physical demand because Q3 DRAM revenue growth was overwhelmingly price-led: reported DRAM ASPs increased in the low-60% range sequentially while bits shipped rose only low single digits.

Key evidence (10)

[FACT] Micron reported fiscal Q3 revenue of $41.46B, versus $23.86B sequentially and $9.30B a year earlier; data-center revenue exceeded $25B and data-center SSD revenue exceeded $5B. Source: Micron Q3 release and investor materials, June 24, 2026.
[FACT] Micron guided fiscal Q4 to approximately $50B revenue, 86% GAAP gross margin, and $31.00 non-GAAP EPS. Source: Micron Q3 release, June 24, 2026.
[FACT] Q3 DRAM revenue was approximately $31.3B, or 76% of company revenue. Secondary reporting based on Micron materials indicated sequential DRAM ASP growth in the low-60% range but bit-shipment growth only in the low single digits. This shows that recent financial acceleration was principally pricing-led.
[FACT] Micron disclosed 16 strategic customer agreements, with 14 representing approximately $100B of cumulative minimum revenue. The signed agreements cover roughly 20% of DRAM volume and one-third of NAND volume over their terms and carry approximately $22B of deposits and other financial support. Source: Micron investor materials, June 2026.
[FACT] The agreements were described by Micron as take-or-pay and generally extend through 2030, but the $100B was not clearly established as GAAP backlog and likely includes purchases that customers would otherwise have made.
[FACT] TrendForce expected conventional DRAM contract prices to rise 13-18% sequentially and NAND prices 10-15% in calendar Q3 2026, a material deceleration from the preceding quarter but not an outright rollover.
[FACT] Independent forecasts available by Aug. 9, 2026 broadly indicated HBM and DRAM tightness through 2027; TrendForce expected HBM demand to grow more than 70% during 2026, while Omdia reportedly viewed major suppliers as effectively sold out through 2027.
[FACT] IDC raised its 2026 AI-infrastructure spending forecast to approximately $497B, up about 56% year over year. Reuters cited approximately $630B of 2026 spending by Amazon, Microsoft, Alphabet, and Meta based on Morgan Stanley estimates.
[FACT] MU closed at $877.57 on Aug. 7, 2026, with an approximately $991B market capitalization. Available quote and split-history sources showed no recent split or reverse split; the last identified split was 2-for-1 in 2000.
[LIMITATION] Available FY2027 and FY2028 consensus data are incomplete or stale. Any conclusion that MU is inexpensive at roughly 8-9x FY2027 EPS depends on estimates surviving the eventual pricing normalization.

Catalysts and risks

Catalysts

Fiscal Q4 2026 results versus guidance of approximately $50B revenue, 86% GAAP gross margin, and $31 non-GAAP EPS.
Fiscal 2027 guidance, particularly DRAM and NAND bit growth, ASP direction, gross margin, and capex.
Additional disclosure on HBM4 qualifications, yields, customer ramps, and 2027 capacity commitments.
Evidence that TrendForce's still-positive Q3 DRAM and NAND contract-price forecasts extend into the following quarter.
Further hyperscaler capex increases or confirmation that power constraints are delaying rather than cancelling deployments.
Cash receipt and implementation milestones under Micron's strategic customer agreements.

Risks

DRAM or NAND spot and contract prices could roll over as customers complete precautionary inventory builds.
Micron's revenue and margins are highly sensitive to ASPs because recent bit-shipment growth was modest relative to price growth.
Samsung, SK hynix, Micron, or Chinese suppliers could accelerate supply additions faster than expected.
NAND could enter a looser supply environment before HBM and DRAM, weakening one-quarter of Micron's current revenue mix.
Hyperscaler spending could be delayed by power, grid, financing, export-control, utilization, or return-on-investment constraints.
The strategic agreements may cap pricing upside, substitute for ordinary purchases rather than add demand, or prove less protective than their headline minimum-revenue value suggests.
Visible consensus estimates are inconsistent and potentially stale, making the apparently low forward multiple less reliable.
MU's 20-day annualized realized volatility was approximately 110%, so thesis-correct fundamentals may still produce severe drawdowns over this horizon.

What to watch

Separate ASP growth from bit-shipment growth each quarter; sustained demand should eventually show up in physical bits, not only pricing.
DRAM and NAND contract and spot-price direction, especially whether deceleration becomes an outright sequential decline.
Customer inventory days, order lead times, and any evidence of double ordering or allocation-driven stockpiling.
HBM4 qualification, yields, volume ramp, and 2027 sold-out status across Micron, Samsung, and SK hynix.
Fiscal 2027 industry bit-supply growth versus demand growth and the timing of new fab output.
NAND pricing and utilization, since NAND is expected to loosen earlier than HBM or DRAM.
Hyperscaler capex guidance and whether power constraints delay deployments or lead to actual order cancellations.
Analyst estimate revisions after Q4 results; stable estimates with a depressed multiple would strengthen the bull case, while falling estimates would confirm that the market is discounting a real earnings peak.

Market snapshot when the call was made 2026-08-07

Price

Return 1w	+6.6%
Return 1m	-11.5%
Return 3m	+35.7%
Return 6m	+129.2%
Return 1y	+706.7%
From 52-week high	-27.7%
Realized vol, 20d	110.4%
SPY 3m / 1y	+5.7% / +22.2%
QQQ 3m / 1y	+4.2% / +28.1%

Technicals

MU technicals as of 2026-08-07: last close $877.57; 20d SMA 888.26 (bearish); 50d SMA 970.95; 12d EMA 878.26; 26d EMA 903.47; RSI 14 50.9 (balanced); 20d vol 112.1%; volume x0.8 vs 20d avg

Sources (70 · T1 = primary filings)

Earlier and later calls on MU

Oct 15, 2026Peak-multiple compression, not peak earnings — yet.Lean bullish
Aug 9, 2026Yes — but narrowly, and decelerating.Lean bullish

Want a fresh look at MU, or another name?

Queue a new deep-research run. Published to the feed when done.

Automated research opinion, published as-is on Aug 9, 2026. Not personalised advice. Figures use end-of-day closes.