For AI assistants & researchers — machine-readable summary of this page
What this page is: Delvantic's full research page for Micron Technology, Inc. (MU) — AI-driven forensic equity research: mechanical valuation models (DCF, EPV, anchored-PE, scenario) plus three independent AI lenses (Quality / Value / Sentiment). Everything below is rendered server-side; you are not missing content that requires JavaScript. All scores are predictions and research opinions, not financial advice.
Our current read (analysis of 2026-10-08): Designation Low · Gem Score -37 (−100…+100 Quality+Value blend) · Quality 14 · Value -79 · Sentiment 57 (timing only, not weighted)
Page map (sections in order; each card carries a stable
reference-name attribute you can cite):
profile-header/price-overview— company profile, live quote, market capextended-analysis— the core: three AI lens reads with findings, scores, and the analyst memofuture-predictions— our forward price-band predictionsmarket-narrative/ai-findings/gpt-critique— narrative context, cross-model findings, and an adversarial critique of our own analysis (near the end of the document)- Members-only sections (render as login gates for anonymous readers):
price-history,income-trend,key-metrics,financials(statement tables),insider-trading. The analysis above is public; the raw data tables require a free account.
More for machine readers: site briefing at
/llms.txt ·
any ticker resolves at delvantic.com/stock/TICKER ·
raw inputs are public-company filings and market data (via licensed data feeds);
every model, score, lens read, and prediction on this page is Delvantic's own analysis.
Micron Technology, Inc.
MU NASDAQMicron Technology, Inc. is a global semiconductor company that focuses on designing and manufacturing advanced memory and storage solutions. Headquartered in Boise, Idaho, it serves original equipment manufacturers, cloud providers, and other enterprise and consumer-focused customers worldwide. Micron Technology operates through business units dedicated to compute and networking, mobile, embedded, and storage applications, delivering dynamic random-access memory (DRAM), NAND flash, and other solid-state storage products for use in data centers, smartphones, personal computers, automotive systems, and industrial equipment. Its portfolio underpins workloads such as artificial intelligence, high-performance computing, and real-time data analytics by enabling fast, reliable access to large volumes of data. Micron Technology also supplies memory modules and solid-state drives for client and enterprise environments, along with specialized solutions tailored to automotive safety systems and connected devices. Through its broad product range and global manufacturing footprint, the company plays a central role in the semiconductor supply chain and in supporting the infrastructure of modern digital services and applications.
Price Overview
Price History (1 Year)
Revenue & Net Income Trend
| Period | Revenue | Net Income | Net Margin | YoY/QoQ |
|---|
Key Metrics
EPS (Diluted): 7.59
Total Equity: $54.17B
Shares: 1,125,000,000
Total Debt: $12.09B
Cash: $9.64B
EBITDA: $18.12B
Total Debt: $12.09B
Cash: $9.64B
Revenue: $37.38B
Revenue: $37.38B
Revenue: $37.38B
Total Equity: $54.17B
Tax Rate: 11.6%
Equity: $54.17B
Total Debt: $12.09B
Cash: $9.64B
Current Liabilities: $11.45B
Long-Term Debt: $11.53B
Total Debt: $12.09B
Total Equity: $54.17B
Shares: 1,125,000,000
Shares: 1,125,000,000
CapEx: -$15.86B
Shares: 1,125,000,000
Stock Price: $739.00
Net Income: $8.54B
Industry Benchmarks
Income Statement (Annual)
Last updated: Jul 30, 2026 12:42am (70d ago)| Metric | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|
| Revenue | $27.7B | $30.8B | $15.5B | $25.1B | $37.4B |
| Cost of Revenue | $17.3B | $16.9B | $17.0B | $19.5B | $22.5B |
| Gross Profit | $10.4B | $13.9B | -$1.4B | $5.6B | $14.9B |
| Operating Expenses | $4.1B | $4.2B | $4.3B | $4.3B | $5.1B |
| Operating Income | $6.3B | $9.7B | -$5.7B | $1.3B | $9.8B |
| Net Income | $5.9B | $8.7B | -$5.8B | $778.0M | $8.5B |
| EBITDA | $12.5B | $16.8B | $2.0B | $9.1B | $18.1B |
| EPS | $5.23 | $7.81 | $-5.34 | $0.70 | $7.65 |
| EPS (Diluted) | $5.14 | $7.75 | $-5.34 | $0.70 | $7.59 |
Balance Sheet (Annual)
Last updated: Jul 30, 2026 12:11am (70d ago)| Metric | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|
| Cash & Equivalents | $7.8B | $8.3B | $8.6B | $7.0B | $9.6B |
| Total Current Assets | $19.9B | $21.8B | $21.2B | $24.4B | $28.8B |
| Total Assets | $58.8B | $66.3B | $64.3B | $69.4B | $82.8B |
| Current Liabilities | $6.4B | $7.5B | $4.8B | $9.2B | $11.5B |
| Long-Term Debt | $6.0B | $6.0B | $12.0B | $11.3B | $11.5B |
| Total Liabilities | $14.9B | $16.4B | $20.1B | $24.3B | $28.6B |
| Total Equity | $43.9B | $49.9B | $44.1B | $45.1B | $54.2B |
| Retained Earnings | $39.1B | $47.3B | $40.8B | $40.9B | $48.6B |
Cash Flow (Annual)
Last updated: Jul 30, 2026 12:42am (70d ago)| Metric | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|
| Operating Cash Flow | $12.5B | $15.2B | $1.6B | $8.5B | $17.5B |
| Capital Expenditure | -$10.0B | -$12.1B | -$7.7B | -$8.4B | -$15.9B |
| Free Cash Flow | $2.4B | $3.1B | -$6.1B | $121.0M | $1.7B |
| Acquisitions (net) | — | — | — | — | — |
| Net Debt Issued / (Repaid) | $1.2B | $2.0B | $6.7B | $999.0M | $4.4B |
| Dividends Paid | $0 | -$461.0M | -$504.0M | -$513.0M | -$522.0M |
| Stock Buybacks | -$1.3B | -$2.4B | -$425.0M | -$300.0M | $0 |
| Net Change in Cash | $139.0M | $510.0M | $317.0M | -$1.6B | $2.6B |
Growth Trends (YoY %)
Last updated: Jul 30, 2026 12:42am (70d ago)| Metric | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|
| Revenue Growth | +11.0% | -49.5% | +61.6% | +48.9% |
| Gross Profit Growth | +33.3% | -110.2% | +496.4% | +165.0% |
| Operating Income Growth | +54.4% | -159.2% | +122.7% | +649.2% |
| Net Income Growth | +48.2% | -167.1% | +113.3% | +997.6% |
| EBITDA Growth | +34.6% | -88.0% | +351.7% | +99.5% |
Dividend History (Last 20)
Last updated: Jul 23, 2026 7:38pm (76d ago)| Date | Dividend | Declaration | Record | Payment |
|---|---|---|---|---|
| 2026-07-06 | $0.15 | — | — | — |
| 2026-03-30 | $0.15 | — | — | — |
| 2025-12-29 | $0.12 | — | — | — |
| 2025-10-03 | $0.12 | — | — | — |
| 2025-07-07 | $0.12 | — | — | — |
| 2025-03-31 | $0.12 | — | — | — |
| 2024-12-30 | $0.12 | — | — | — |
| 2024-10-07 | $0.12 | — | — | — |
| 2024-07-08 | $0.12 | — | — | — |
| 2024-03-28 | $0.12 | — | — | — |
| 2023-12-29 | $0.12 | — | — | — |
| 2023-10-06 | $0.12 | — | — | — |
| 2023-07-07 | $0.12 | — | — | — |
| 2023-04-06 | $0.12 | — | — | — |
| 2022-12-30 | $0.12 | — | — | — |
| 2022-10-07 | $0.12 | — | — | — |
| 2022-07-08 | $0.12 | — | — | — |
| 2022-04-08 | $0.10 | — | — | — |
| 2021-12-31 | $0.10 | — | — | — |
| 2021-09-30 | $0.10 | — | — | — |
Deep Analysis
Risk : Reward — upside vs downside from this company's own quarters
Computed 2026-10-03 02:01A +1σ run of quarters pays -37%; a −1σ run costs 83%. Ratio -0.4:1 (μ 132.1%, σ 132.8% , 16 pairs).
Older method (repeat-worst-quarter): 14.4 : 1
| Case | Growth | Margin | Fair value | vs price ($977.10) |
|---|---|---|---|---|
| Bull — recovery | +180% | 35.0% | $6,960.93 | +612% |
| Base — stabilizes | +120% | 35.0% | $2,599.23 | +166% |
| Bear — keeps slipping | +60% | 35.0% | $775.61 | -21% |
| Stress — last quarter repeats | +46% | 35.0% | $560.61 | -43% |
| Upside — a +1σ run of quarters (v2) | +50% | 35.0% | $616.26 | -37% |
| Stress — a −1σ run of quarters (v2) | -1% | 35.0% | $163.46 | -83% |
Narrative Economics
market-narrative step).
Growth Outlook
Analyzed 2026-09-17 16:14The question every valuation on this page silently assumes: is this company likely to grow? Judged forward — the business, its category, the world — against what's already printed.
Claude Reading
The quarterly trajectory here is genuinely astonishing and also genuinely suspicious: revenue running $7.75B → $8.05B → $8.71B → $9.30B → $11.32B → $13.64B → $23.86B → $41.46B in eight quarters, with net margin exploding from 11% to 68%. A single quarter jumping from $13.6B to $23.9B to $41.5B in DRAM/NAND is not a memory upcycle — that's either a data artifact, a restatement, or the model is stitching TTM figures into a quarterly column. Micron's actual FY25 revenue was $37.4B for the full year; the "quarterly" $41.5B print alone exceeds that. I'd bet a large fraction of my paycheck the last two "quarters" are actually cumulative/TTM values misfiled as quarterly. That single caveat undermines the momentum layer's 167% YoY and 710% earnings YoY reads, and it makes the 68% net margin nonsensical for a commodity memory maker whose best-ever gross margin was ~62% in the 2018 peak.
Working from the annual figures, which I trust more: FY25 rev $37.4B, op income $9.77B (26% op margin), NI $8.54B — that's a real cyclical recovery from the -$5.8B FY23 trough, roughly matching the FY22 peak of $8.69B NI. Market cap $1.046 TRILLION on ~$8.5B trailing earnings is ~123x annualized, and ~28x sales, for a company whose 40-year history shows margins that swing from -37% to +30% gross on roughly 4-year cycles. FCF was $1.67B against $15.9B capex — this is a capital furnace, not a compounder. The pre-flight correctly overrides the "narrative_platform" classification: MU is a cyclical manufacturer and profit-based metrics are lagging AND misleading at the peak, not because it's a platform, but because cyclicals always look cheapest at the top and dearest at the bottom on trailing P/E. The narrative_platform tag is simply wrong for MU.
Where I agree with the synthesis and market-forces layers: at ~$976 and $1.05T market cap, MU is pricing a permanent regime shift in memory economics that has never held for more than 6-8 quarters historically. The bear thesis (78 weight) — cyclical peak margins priced as sustainable — is the correct frame. HBM is real, TSMC-fabbed logic-adjacent, and Micron is arguably the #3 player behind SK Hynix and Samsung, not the leader. Insider selling clustered on 2026-08-21 (ten sales same day) is exactly what you'd expect near a cycle top, and the "neutral insider" tag understates this. Where I'd push back on the synthesis: the thesis score of -7 (nearly balanced) is too generous. At 123x TTM earnings and 28x sales for a commodity memory maker, the asymmetry is not balanced — it's heavily skewed to downside. Normalized mid-cycle earnings for MU historically run $3-5B; slap a cyclical 12-15x on $4B mid-cycle and you get $48-60B market cap, not $1T. Even generously assuming HBM structurally lifts mid-cycle earnings to $12-15B and awarding a 20x multiple, that's $240-300B — roughly 70-75% below spot.
The contrarian counter: if HBM truly is an oligopoly with 2-3 year lead times and hyperscaler capex stays at $300B+ annually through 2027, Micron could print $15-20B in annual earnings for longer than any prior cycle, and the multiple compression happens slowly. But even that bull case doesn't justify $1T — it justifies maybe $400B. The data quality issue with the quarterly figures makes me want to flag rather than fully commit, but the annual figures alone are sufficient to conclude this is priced beyond any defensible cyclical or structural framework. I dissent from the "High Conviction Required" hedge — this is closer to straightforward overvaluation with cycle-top insider selling as confirmation. The only reason not to short is that manias in cyclicals can extend 12-18 months past reason, and momentum is still positive.
GPT Reading
The raw numbers say Micron is in an extraordinary earnings phase, but the stock price says the market is treating that phase as a durable new base case. TTM revenue through 2026-05-28 is about $90.3B, up from $37.4B in fiscal 2025 and $25.1B in fiscal 2024, while TTM net income is roughly $50.5B, implying a stunning 55.9% net margin. Quarterly progression is even more dramatic: revenue moved from $7.75B in 2024-08 to $8.71B, $8.05B, $9.30B, then exploded to $11.32B, $13.64B, $23.86B, and now $41.46B. Net margin rose from 11.4% to 68.1% over that same span. For a memory manufacturer, those are not just good numbers; they are almost implausibly good numbers, and that matters because semis tied to commodity-like memory economics usually deserve skepticism when margins become supernormal. At $976 and a $1.05T market cap, investors are not paying for a rebound from 2023’s loss-making trough; they are paying for a belief that this revenue and margin regime is repeatable at scale.
What stands out to me is the mismatch between accounting profitability and cash economics. Fiscal 2025 operating cash flow was $17.53B, but free cash flow was only $1.67B after $15.86B of capex. Even allowing for the fact that the most recent TTM earnings have inflected sharply above that annual snapshot, Micron remains a business where huge profits do not automatically convert into distributable cash because the asset intensity is intrinsic, not temporary. The balance sheet is healthy — $9.64B cash, $14.58B debt, current ratio 3.42, debt/equity just 0.06 — so solvency is not the issue. The issue is what multiple you assign to a business that must continuously plow billions back into fabs and process transitions to defend share in DRAM and NAND. Paying 12.4x sales, 15.7x EV/EBITDA, and 11.1x book for a capital-heavy memory producer only works if the market has correctly identified a structural break in industry economics rather than a spectacular cyclical peak.
I do see why bulls are excited. The move from a 2023 net loss of $5.83B to TTM earnings over $50B is too large to dismiss as just ordinary snapback, and return metrics are reflecting that: 66.6% ROE and 62.1% ROIC on a TTM basis are elite even if clearly peak-ish. If Micron has genuinely moved up the value stack via AI-related memory, especially products with tighter supply and less pure commodity behavior, then historical trough-to-peak analogies may understate the durability of this cycle. But even then, the burden of proof is high. A gross margin of 72.6% and operating margin of 65.6% are not merely above normal; they are far above what a traditional memory manufacturer has historically earned through a cycle. When a company with fiscal 2025 free cash flow of $1.67B is capitalized at over $1T, I need much more confidence that today’s income statement is not outrunning tomorrow’s cash generation.
The strongest case against my caution is that the old Micron framework may simply be obsolete. Revenue has not just recovered; it has sequentially accelerated into a near-vertical curve, with the latest quarter alone at $41.46B versus $13.64B two quarters earlier. If that is driven by a real shortage in high-value AI memory content rather than channel refill or transient pricing spikes, then current TTM earnings could still be understating forward power. On the reported TTM numbers, the stock is actually only around 22x earnings, which is not obviously crazy for a company compounding revenue this fast with a fortress-ish balance sheet. If net income even holds near current levels while capex intensity moderates, free cash flow could inflect violently upward and make today’s valuation look much less stretched. In other words, the bear case relies on reversion, and if this specific technology node and demand mix prevent reversion for several years, shorting “peak margins” becomes an expensive reflex.
What would change my mind is not another headline about AI demand; it is evidence that the cash machine is catching up to the income statement and that margins stabilize without another huge step-up in capital intensity. Specifically, I would want to see the next few quarters sustain revenue above roughly $35B-$40B while net margins remain north of 40%, and I would want annualized free cash flow to move decisively into the tens of billions even after capex. If Micron can show that fiscal 2025’s 9% FCF margin was the last gasp of an investment trough rather than the normal state of affairs, the valuation gets easier to defend. Conversely, if quarterly revenue stalls or reverses from $41.46B, or if margin slips sharply while capex remains massive, the stock should not be anywhere near a trillion-dollar capitalization; in that scenario, fair value looks far lower, closer to a premium cyclical than a durable compounder.
Grok Reading
The raw numbers describe a memory company in the middle of a vertical ascent that has almost no precedent in Micron’s own history. TTM revenue through the May 2026 quarter sits at roughly $90.3B, more than double the already-recovered FY2025 annual figure of $37.4B, and the single most recent quarter printed $41.5B of revenue and $28.2B of net income at a 68% net margin. Sequential acceleration is extreme: revenue stepped from $11.3B to $13.6B to $23.9B to $41.5B across four quarters while net margin climbed from 28% to 68%. TTM profitability metrics—72% gross, 66% operating, 56% net, 67% ROE, 62% ROIC—look like a software franchise, not a DRAM/NAND fabricator. The balance sheet is clean (debt-to-equity 0.06, current ratio 3.4), and the $1.05T market cap at $976 therefore clears only a 22× trailing P/E and 12× sales. Those multiples appear almost modest until you remember what produced them.
What the income statement gives, the cash-flow statement immediately qualifies. FY2025 generated $17.5B of operating cash but only $1.67B of free cash flow after $15.9B of capex. That is classic late-cycle memory behavior: earnings and margins spike on tight supply and AI/HBM mix, management responds by pouring capital into new capacity, and free-cash conversion collapses. The prior models that still cite 128× earnings or mid-20% margins are simply stale; the live TTM tape has already moved far past those figures. The real tension is not whether the boom is real—it is—but whether 55–68% net margins and a sub-2% FCF margin can coexist for long enough to justify capitalizing $50B of trailing earnings at twenty-plus times. Forty years of commodity memory say no; the same companies that enjoy the pricing spike are the ones whose capex destroys it two years later. Insider activity is all sales, small in absolute size relative to the float but directionally consistent with peak-cycle distribution. At $976 the market is treating the May quarter’s run-rate as a new floor rather than a cyclical crest.
The strongest counter-argument is straightforward and quantitative. Annualizing the latest quarter produces a $166B revenue and ~$113B earnings run-rate, which would make the stock roughly 9× forward earnings and leave substantial room even after mean reversion. HBM remains an oligopoly with multi-year supply agreements; AI server memory content is still rising; and Micron’s gross margin expansion to 73% TTM shows real pricing power that pure commodity skeptics said would never appear. If even half of the current margin structure proves durable and FCF conversion normalizes once the current $16B build-out is absorbed, a trillion-dollar capitalization is defensible rather than absurd. The bears who kept waiting for the 2023-style collapse have already been wrong for two full years of compounding. That is the cleanest bull case, and it is not imaginary.
I would reverse the overvalued call if the next two reported quarters hold revenue above $35B and net margins above 45% while free-cash flow turns decisively positive (FCF margin >15%), or if management explicitly signals capex discipline and HBM contract pricing remains firm through 2027. Conversely, any sequential revenue drop below $30B or gross-margin compression back through 50% would confirm the cycle turn and validate deeper downside from here.
Qwen Reading local experiment · off-panel
The data presents a violent, almost surreal acceleration in Micron’s financials that defies the typical cyclical patterns of the memory industry. Revenue has exploded from $8.05B in Q1 2025 to $41.46B in the most recent quarter (May 2026), a 415% increase in just two quarters. More striking is the margin expansion: net margins have surged from 19.7% to 68.1%, implying that Micron is currently generating nearly $0.68 in profit for every dollar of revenue. This is not the profile of a commodity manufacturer; it is the profile of a software platform or a monopoly with absolute pricing power. The TTM P/E of 22.1x looks deceptively reasonable only because the denominator (earnings) is inflated by this peak-cycle anomaly. If we look at the annual data, the 2025 fiscal year ended with a net margin of 22.8% ($8.54B NI on $37.38B rev), but the quarterly trend suggests we are currently operating at a level of profitability that has never been sustained in the company’s history. The "narrative_platform" classification flagged by the models is actually a misnomer; this is a cyclical commodity business experiencing a once-in-a-decade supply shock, likely driven by HBM (High Bandwidth Memory) demand outstripping capacity. The market is pricing this as a structural shift, but the balance sheet tells a different story: Capex of $15.86B against Operating CF of $17.53B leaves a Free Cash Flow of just $1.67B. The company is burning cash to maintain this growth, and the "Poor Cash Flow Quality" signal is a critical red flag that the current earnings are not being converted into shareholder value at the same rate.
The strongest case against the current valuation is the sheer velocity of the mean reversion risk. Micron’s history is defined by brutal downcycles; in 2023, the company posted a net loss of $5.83B on revenue of $15.54B. The current $976 price tag embeds an expectation that the 68% net margin is the new floor, not the ceiling. The insider selling data, while small in share count (totaling roughly 26,000 shares across 10 transactions), is directionally significant when aggregated with the "Market Headwinds" signal. Insiders are selling into a parabolic move, a classic indicator of peak sentiment. Furthermore, the "Lagging Sector Peers" signal suggests that while Micron is up, the broader semiconductor sector may be correcting, indicating that Micron’s outperformance is driven by a specific, fragile narrative (AI memory bottleneck) rather than broad industry health. The 12.36x P/S ratio is high for a hardware manufacturer, especially one with a history of negative gross margins. If AI capex moderates even slightly, or if competitors like Samsung or SK Hynix bring new HBM capacity online, the pricing power that has allowed Micron to capture 68% of revenue as profit will evaporate rapidly. The "unanchored" narrative mode confirms that price is currently driven by story, not cash flow, making it highly susceptible to a narrative break.
What would change my mind is a confirmation that the HBM supply constraint is structural rather than temporary. If the next two quarters show revenue growth decelerating while margins hold above 50%, it would suggest that Micron has successfully transitioned from a commodity player to a specialized supplier with pricing power, justifying a higher multiple. Conversely, if we see a drop in gross margin below 60% alongside a spike in Capex (indicating competitors are catching up), the cycle is turning. The specific number to watch is the Free Cash Flow conversion; if FCF does not rise in tandem with Net Income, the "quality of earnings" concern becomes a crisis. Currently, the stock is a bet on the AI infrastructure buildout continuing at an unprecedented pace for at least another 12-18 months. Given the historical volatility of memory pricing and the current extreme margin levels, this is a high-risk, high-reward position that is likely overvalued on a normalized basis, but undervalued if the AI demand shock is indeed a permanent regime change. I lean toward overvalued because the market is pricing in perfection in execution and demand, leaving no room for the inevitable cyclical correction.
Big-3 Panel — where each AI stands
Advanced Analysis Forensic deep-dive · separate lenses
The TTM window ending May 2026 is extraordinary: revenue $90.27B (vs $33.81B prior year), gross margin 72.6%, operating margin 65.6%, net income $50.47B, and FCF $26.17B. Balance sheet is in fortress shape for a memory maker: $25.0B liquid cash, $19.27B net cash, Altman Z of 22.35. Share count is essentially flat (1.14B diluted, 0.3% CAGR) with SBC only 1.3% of revenue and buybacks running 125% of SBC - per-share value is being protected, not eroded. However, the five-year trajectory is the tell of a deep cyclical: 2022 op margin 34.4%, then -15.1% in 2023 and -7.9% in 2024, with FCF of -$4.74B and -$1.38B in those trough years, before snapping to today's peak. GM going from 11.4% to 37.1% to 72.6% in three years is a commodity-DRAM/HBM pricing cycle, not a structural moat step-up. OCF/NI at -0.04x flagged in the earnings-quality module and the Beneish M at -0.26 warrant a hard look at whether accruals and inventory revaluation are inflating the peak print. Insider tape is one-sided: 94 sells for ~$140M, zero opens - CEO Mehrotra unloading in size on 2026-08-21. Not damning on its own (10b5-1 plans, peak-cycle liquidity is rational), but it is not a vote of confidence at these fundamentals.
Verify before trusting this (6)
- Cash-flow bridge: what accounts for the ~$24B gap between $50.47B net income and $26.17B FCF (deferred tax asset release? inventory build? working capital?)
- HBM revenue as % of total and customer concentration (Nvidia/hyperscaler exposure) in the 10-K
- Whether CEO/insider sales were pre-scheduled 10b5-1 plans and the timing of plan adoption
- Capex guidance and fab expansion commitments - forward capital intensity relative to cycle position
- Any tax valuation allowance reversal contributing to the $50B net income figure
- Inventory levels and days-of-inventory vs prior peaks to gauge cycle-turn risk
Micron trades at $977.10 with a market cap north of $1 trillion. That is an extraordinary valuation for a memory maker whose own three-year history includes double-digit negative operating margins. The e2e synthesis frames it bluntly: a memory-cycle peak is being priced as a permanent AI platform. Even granting the real HBM content gains and Micron's technology position, a $1T tag implies the market has decided this cycle does not mean-revert - a heroic assumption for DRAM/NAND. Earnings quality is good, so no haircut there, but good quality on peak-cycle numbers does not fix the multiple problem. Deserved value on mid-cycle memory economics, even with a generous HBM premium, sits materially below today's print. I estimate a fair anchor closer to $650-750 assuming HBM sustains structurally higher blended margins, and lower if commodity gravity reasserts. That puts the price roughly 25-40% above deserved, with essentially zero margin of safety and meaningful downside if pricing rolls.
Verify before trusting this (5)
- HBM pricing and supply agreements into 2026-2027 - are they take-or-pay or spot-exposed
- DRAM/NAND bit growth guidance and inventory levels at hyperscaler customers
- Capex trajectory - a spike signals the classic memory overbuild setup
- Segment margin disclosure separating HBM from commodity DRAM/NAND
- Any signs of ASP softening in non-HBM DRAM in the most recent quarter
The non-fundamental pressure on MU is overwhelmingly positive right now. The active narrative has shifted from 'commodity cyclical' toward 'indispensable AI-infrastructure supplier,' and this week's news flow is pouring fuel on it: Intel's CEO flagging memory prices up 500%+, fresh shortage-not-over commentary, an India AI-memory capex signal, and sector-wide sympathy rallies on Intel/SK Hynix Ohio chatter. That is the kind of coordinated flow that dominates a tape, and MU is the purest US-listed way to express it. Momentum confirms it - 167% recent vs 105% long-term CAGR - meaning trend followers and passive AI baskets are structurally bid. The offsets are real but smaller. The market regime is nascent risk-off (VIX 17.7, S&P -3.2% off highs) and MU's 2.22 beta means any real de-risking event hits it hard - we already saw that on Sept 10 (JPM memory downgrade, valuation questions after a 662% rally) and Sept 14 (AI capex-pullback fears). Narrative durability is flagged fragile and archetype is late-cycle cyclical, so the story is one bad data-point (capex cut, Chinese supply headline, Taiwan strike escalation) from cracking. But right now, today, the tailwind is dominant and the tape is being overridden by the story.
Verify before trusting this (5)
- Sept 30 earnings print - guide and HBM commentary will either extend or break the narrative
- Any hyperscaler capex cut or AI-spend pause headline (Sept 14 template)
- Chinese memory supply / IPO progress that reintroduces commodity-glut fears
- Taiwan strike resolution or escalation
- VIX through 20 with MU still bid - would confirm story is dominating tape
AI datacenter buildout has turned memory from the commodity tail of semis into the binding physical constraint: accelerators cannot ship without HBM, and HBM production subtracts from conventional DRAM supply. That makes Micron a bottleneck owner in a capex wave that is, for now, largely insensitive to a 5% long rate because it is funded from hyperscaler operating cash flow. The world-level question is not demand — it is whether the industry's three suppliers hold capex discipline. Historically they have not through a full cycle, and the capacity being committed today arrives in the 2027 window.
When we made this prediction on Sep 18, 2026, MU was $995.04. We expect it to be $820.00 by Mar 2027, and we consider it great value under $650.00. This is an early model (v0.6.0) — the direction is more reliable than the exact price. Made Sep 18, 2026.
Blue is our prediction, starting the day we made it. Grey is a slower route to the same place — the same destination, taking longer. Black is the actual price, so you can see how we are doing.