Micron – why the short thesis may be too simple.
First of all, Micron is normally treated by investors as a classic cyclical memory company. In a normal cycle that would be a good starting point: DRAM and NAND prices rise, producers add capacity, prices collapse, margins disappear, and then the cycle repeats. This is the historical reason why some investors are tempted to short Micron after a huge rally.
But this time the situation is not only about a normal memory cycle. Micron is now benefiting from AI data centers, HBM, high-capacity server DRAM, data center SSDs, automotive memory, industrial applications, mobile memory and PC refresh. The company also signed strategic customer agreements that can change the volatility profile of the business. In my view, this is the key argument against a pure cyclical short thesis.
Michael Burry reportedly disclosed a short position against Micron around $1,051.87 per share, arguing that AI enthusiasm, FOMO and semiconductor overinvestment may have pushed the stock into bubble territory. That argument is understandable after a huge price move. However, the counterargument is that Micron's financial profile is no longer only spot pricing plus cyclical inventory. The company now has customer commitments, pricing floors in some contracts, customer deposits and demand from several end markets.
Main thesis: Burry may be right about valuation risk and market overheating, but he may be underestimating the durability of Micron's demand base. The short can work if memory prices collapse or AI capex slows sharply. But if DRAM/NAND tightness persists through 2027 and SCAs stabilize pricing, Micron may remain financially stronger for longer than historical cycles suggest.
Latest quarterly situation – FQ3 2026
Micron reported a record fiscal Q3 2026. Revenue was $41.46 billion, up 74% sequentially and 346% year over year. GAAP net income was $28.24 billion and operating cash flow was $25.39 billion. The company guided fiscal Q4 revenue to around $50.0 billion plus or minus $1.0 billion, with very high gross margin expectations. This means that, based on Q1–Q3 actual results and Q4 guidance, fiscal 2026 revenue could be around $129 billion.
| Quarter / item | Revenue | Operating cash flow | Comment |
|---|---|---|---|
| FQ1 2026 | $13.64B | $8.41B | Strong start of the AI-memory acceleration; tight DRAM supply and HBM demand. |
| FQ2 2026 | $23.86B | $11.90B | Revenue nearly tripled year over year; gross margin reached around 75%. |
| FQ3 2026 | $41.46B | $25.39B | Record quarter; DRAM and NAND pricing increased sharply. |
| FQ4 2026 guidance | $50.0B ± $1.0B | FCF expected to increase substantially | Management expects another step-up in revenue and cash generation. |
| FY2026 estimate | ~$129B | Very strong; exact full-year figure pending Q4 | Based on reported Q1–Q3 plus Q4 company guidance. |
Annual financials – last 3 fiscal years
Before the 2026 boom, Micron already showed a sharp recovery from the 2023 memory downturn. Revenue rose from $15.54 billion in FY2023 to $37.38 billion in FY2025. EBIT moved from a deep operating loss to $9.77 billion of operating income. Operating cash flow moved from $1.56 billion to $17.53 billion.
| Fiscal year | Revenue | EBITDA | Operating income / EBIT | Operating cash flow | Adjusted free cash flow |
|---|---|---|---|---|---|
| FY2023 | $15.54B | ~$2.11B | -$5.75B | $1.56B | Negative |
| FY2024 | $25.11B | ~$9.08B | $1.30B | $8.51B | Approx. negative / low because capex remained high |
| FY2025 | $37.38B | ~$18.12B | $9.77B | $17.53B | $3.72B |
Revenue acceleration – FY2025 vs FY2026
In our previous discussion, the key question was whether Micron's revenue is going to multiply by four. The corrected answer is: based on current reported quarters and company guidance, FY2026 revenue is closer to 3.4–3.5 times FY2025 revenue, not exactly 4 times. Still, this is an extraordinary acceleration for a company of this size.
| Year | Revenue | Change | Multiple vs FY2025 |
|---|---|---|---|
| FY2025 actual | $37.38B | Base year | 1.0x |
| FY2026 estimate | ~$129B | ~+245% | ~3.45x |
Why volumes and revenue are rising
The most important point is that Micron's revenue growth is not only volume growth. In FQ3 2026, DRAM revenue increased 67% sequentially, while bit shipments increased only in the low-single-digit percentage range. Prices increased in the low-60s percentage range. NAND revenue increased 99% sequentially, while bit shipments increased in the mid-single-digit range and prices increased in the mid-80s percentage range. In plain English: the cycle is being driven mainly by price and mix, not only by more bits shipped.
| Segment | FQ3 2026 revenue | Sequential revenue growth | Bit shipment trend | Price trend | Interpretation |
|---|---|---|---|---|---|
| DRAM | $31.3B | +67% | Low-single-digit increase | Low-60s % increase | Mostly pricing and mix, helped by HBM and data center demand. |
| NAND | $9.9B | +99% | Mid-single-digit increase | Mid-80s % increase | Extreme tightness in NAND and favorable product mix. |
How long can the volume / pricing cycle last?
Micron management expects tight supply-demand conditions for both DRAM and NAND to persist beyond calendar 2027. In DRAM, industry bit shipments in calendar 2026 are expected to grow in the low- to mid-20% range. In NAND, industry bit shipments are expected to grow approximately 20%. Micron expects its DRAM supply growth to be approximately in line with industry supply growth and its NAND supply growth to be somewhat lower than industry supply growth.
| Market | 2026 industry bit shipment outlook | Micron supply outlook | Supply-demand comment |
|---|---|---|---|
| DRAM | Low- to mid-20% growth | Approximately in line with industry | HBM trade ratio and AI demand pressure non-HBM supply. |
| NAND | Approximately 20% growth | Somewhat below industry | Limited cleanroom space and DRAM/HBM prioritization constrain supply. |
| Overall | Growth but not enough | Disciplined expansion | Management says tightness can persist beyond calendar 2027. |
Backlog / Strategic Customer Agreements for 2026 and 2027
Micron does not have a classic industrial backlog like a machinery company. But the new Strategic Customer Agreements are economically similar to a backlog because they include take-or-pay commitments, committed volumes, customer deposits, and in many cases pricing floors or price bands. This is a major change in Micron's business model.
| Backlog / SCA item | Micron disclosure | Why it matters |
|---|---|---|
| Signed SCAs | 16 strategic customer agreements | Broad customer commitment across data center, consumer, auto and industrial markets. |
| Typical term | Usually calendar 2026 through calendar 2030; automotive often 3 years | Improves visibility into 2026 and 2027 demand. |
| Volume coverage | Roughly 20% of DRAM volume and one-third of NAND volume over the agreement period | Large enough to matter for business stability. |
| Future revenue under SCAs | Management expects approximately half or more of company revenue to be under SCAs when completed | This can reduce cyclicality versus historical Micron. |
| RPO | Over $5B at FQ3 end; approximately $100B including signed SCAs after quarter-end | Important visibility but conservative because it is based on minimum volumes and minimum prices. |
| Cash deposits and commitments | ~$22B projected; roughly $18B in cash deposits | Customers are financially committing to supply, not just giving soft demand forecasts. |
Peer comparison – revenue and 2026 forecast
The table below is a market-context comparison. Micron is the most dramatic growth story because its 2025 base was still early in the recovery and 2026 is being transformed by DRAM/HBM/NAND price strength. Samsung is a broader electronics conglomerate, so memory strength is diluted by other segments. SK Hynix is the closest HBM peer. Western Digital and Kioxia are more NAND/storage exposed.
| Company | 2025 revenue | 2026 revenue forecast / estimate | Growth | Comment |
|---|---|---|---|---|
| Micron | $37.4B | ~$129B | ~+245% | Based on Q1–Q3 actuals plus Q4 guidance; strongest acceleration among memory peers. |
| SK Hynix | ~$92B | ~$155–165B | ~+70–80% | Direct HBM leader; very strong but already larger base. |
| Samsung Electronics | ~$226B | ~$295–305B | ~+30–35% | Memory rebound is diluted by consumer electronics, foundry and other divisions. |
| Kioxia | ~$12B | ~$16–17B | ~+35–40% | NAND recovery, but less HBM exposure than DRAM leaders. |
| Western Digital | ~$15B | ~$19–20B | ~+30% | Storage/HDD/NAND exposure; less direct AI-memory leverage. |
Market multiples and valuation context
After the rally, Micron is no longer cheap on trailing multiples. The market is paying for the possibility that 2026–2027 earnings are not a one-quarter spike but a structurally higher profitability regime. This is exactly where the long vs short debate sits.
| Company | Segment exposure | P/E | P/S | EV/Sales | EV/EBITDA | Valuation comment |
|---|---|---|---|---|---|---|
| Micron / MU | DRAM, NAND, HBM, data center SSD, auto/industrial | ~22.7x | ~12x | ~12x | ~16x | Expensive versus old-cycle Micron, but potentially fair if SCAs and high margins persist. |
| SK Hynix | DRAM, HBM, NAND | ~22x | ~13x | ~13x | ~16x | Closest peer for HBM; valuation broadly comparable to Micron. |
| Samsung Electronics | Memory, foundry, mobile, consumer electronics | ~11x | ~5x | ~5x | ~12x | Cheaper because it is less pure-play memory and has conglomerate discount. |
| Western Digital | Storage / HDD / NAND | ~34x | ~2x | ~17x | ~52x | Trailing EBITDA distorted by cycle; not a perfect Micron peer. |
Why Michael Burry may be wrong on Micron
Let me be clear: shorting Micron after a very large rally is not irrational. Memory stocks are historically cyclical and can collapse when prices peak. Burry's argument may be that the market is extrapolating peak earnings too far into the future. This is a valid risk.
However, there are four reasons why this short may be dangerous:
| Reason | Why it challenges the short thesis |
|---|---|
| 1. Demand is broader than AI only | Micron sells into AI data center, general data center, enterprise SSD, automotive, industrial, mobile and PC. AI is the accelerator, but not the only source of demand. |
| 2. SCAs reduce cyclicality | Take-or-pay agreements, deposits and pricing floors can make revenue and margins more predictable than in past memory cycles. |
| 3. Supply is structurally constrained | Cleanroom constraints, long fab construction cycles, HBM trade ratio, EUV complexity and energy infrastructure limit quick supply response. |
| 4. Cash flow is real | Micron is generating very large operating cash flow and free cash flow. A short thesis based only on valuation must also explain why cash generation will collapse quickly. |
Why Burry may still be right
The bear case should not be ignored. Micron can still be overvalued if the current margins are peak margins, if customers overordered memory, if hyperscaler AI capex slows, or if Samsung/SK Hynix/Micron/Kioxia add too much capacity into 2028. The market is already pricing in a lot of success.
| Risk | Impact on Micron |
|---|---|
| Memory price reversal | Would hit gross margin, EBITDA, earnings and cash flow very quickly. |
| AI capex slowdown | Could reduce HBM demand and weaken pricing power. |
| New capacity in 2028+ | Could normalize supply-demand conditions and reduce scarcity premiums. |
| Valuation compression | Even if profits stay high, P/E and EV/EBITDA multiples can fall if the market fears peak earnings. |
Base case, bull case and bear case
| Scenario | Assumption | Financial implication | Investment interpretation |
|---|---|---|---|
| Bull case | AI/HBM demand remains extremely strong, SCAs protect pricing, supply remains tight beyond 2027. | FY2026–FY2027 revenue and cash flow stay near record levels. | Short thesis fails; Micron deserves a higher structural valuation. |
| Base case | 2026 is exceptional, 2027 remains strong, 2028 gradually normalizes. | Revenue and margins decline from peak but remain far above 2023–2025 trough levels. | Stock may be volatile, but business quality is better than old-cycle Micron. |
| Bear case | Memory pricing turns quickly, AI demand slows, new capacity arrives faster than expected. | EBITDA and FCF fall sharply; market re-rates Micron as a peak-cycle semiconductor name. | Burry-style short can work. |
Foxorox conclusion
Micron is not a risk-free long. The stock has moved violently, valuation is demanding, and memory cycles have historically punished investors who bought peak earnings. But the short thesis is also not simple. Micron now has a combination of AI demand, HBM scarcity, NAND recovery, data center SSD growth, automotive/industrial demand and multi-year strategic customer agreements. This is a more stable business model than the old spot-memory cycle.
My conclusion about its short is as follow: Michael Burry may be right that Micron's stock is vulnerable to a correction after a huge rally. The chart is asking about to short. But the huge increase of prices and demand and encouraging sales is against the lower valuation. Also cashflow looking good. It might be falsify by some opeartions beetwen large moves of invoices as Nvidia is doing through its subsidaries but I assume that Micron demand is more natural that demand for Nvidia;ss Chip. He may be wrong if he assumes that Micron is still just the same old cyclical DRAM company. The company has visible demand into 2026 and 2027, and its customers are signing commitments because memory has become a strategic bottleneck as it wwould be use not only for data centers. In this setup, shorting Micron is not only a bet against valuation. It is a bet against the cycle but cycle is not finished already.
Written by Pawel Demczuk