Micron Technology, Inc.
Micron is riding a multiyear AI memory shortage with locked-in demand, rising earnings power, and a still-below-peer P/E, even after a huge move.
Micron is no longer just a cyclical memory name waiting for the next pricing turn. It is one of the few scaled suppliers at the center of the AI memory bottleneck, and management is backing that statement with numbers that matter. The strongest proof is not tone from the call. It is 16 strategic customer agreements, more than $22B of commitments, nearly $18B of cash deposits, and terms that are explicitly take-or-pay with no cancellation provision. That structure is unusual for memory and meaningfully changes revenue visibility.
The market still does not seem to fully price how different forward earnings power could look if this demand picture persists through 2027 and 2028. FY2025 EPS was $7.65, but the quarterly cadence has exploded higher with $3.03, $4.78, $12.2, and $25.11 over the last four reported quarters. Even allowing for noise in quarterly memory economics, that is the signature of a business moving into a new earnings band. On that basis, 24.11x P/E is not demanding versus a 33.6x peer average. The stock is expensive on 13.5x P/S, but that matters less if margins and mix keep shifting toward HBM and data center SSD.
The case depends on three things. First, AI memory tightness must remain real, and management is saying it does through 2028. Second, Micron must execute the capacity build without destroying returns. FY2026 CapEx is moving to roughly $27B, FY2027 goes higher, and startup costs will ding margins by 0.5 to 1+ points with $100M to $200M quarterly cost burden. Third, customer concentration and hyperscaler urgency have to keep supporting disciplined pricing rather than encouraging oversupply.
What keeps conviction at Medium instead of High is the stock, not the business. MU has rallied 63.7% above its 200-day average and sits near its highs, so near-term expectations are elevated. Still, the underlying business looks headed toward a larger, less purely cyclical, AI-linked earnings stream than the old Micron playbook assumed. That remains worth owning.
HBM supply remains sold out deeper into 2028, enterprise SSD share gains continue, and earnings power closes toward the higher-end multiple framework.
Micron sustains above-cycle earnings growth, converts SCAs into revenue visibility, and absorbs startup costs without breaking the AI memory thesis.
AI demand stays strong but margin compression from FY2027 startup costs, elevated CapEx, and a de-rating from a stock near highs pulls valuation closer to the DCF and P/S cases.
Quant call aligns with analyst consensus (Strong Buy: 22 buy / 1 hold / 0 sell).
Go for a business that any idiot can run — because sooner or later, any idiot probably will.
Forward earnings power is the whole case. Trailing FY2025 EPS of $7.65 puts the stock at 24.11x, but recent quarterly EPS of $25.11, $12.2, $4.78, and $3.03 shows how fast normalized earnings are scaling as AI memory mix improves, even before the next wave of capacity lands. The moat is stronger than it was 3 to 5 years ago because AI memory tightness, packaging complexity, and customer qualification cycles favor scaled incumbents, while Micron now has take-or-pay contracts that reduce classic memory-cycle volatility. AI is a direct tailwind here, not an indirect theme: HBM TAM is now expected to easily cross $100B in 2027, and Micron is one of very few suppliers that matter. The constraint is not demand but supply and capital, which supports a BUY despite near-term cost pressure.
- 16 SCAs with $22B+ commitments, including nearly $18B of cash deposits, lock in multiyear AI memory demand
- FY2025 revenue reached $37.4B, up 46.6% YoY, while operating margin expanded to 19.7% from 5.2%
- Valuation remains 24.11x P/E versus peers at 33.6x, despite eight straight EPS beats including +19.7% and +32.8% in the last two quarters
- AI demand is not theoretical anymore. Management said, "Our demand for HBM is well above our ability to supply not just in 2027 but even 2028, across all HBM flavors. The same is also true for non-HBM DRAM."
- Customer commitment quality improved materially. 16 signed SCAs are take-or-pay, generally 5-year, with quarterly pricing bands and no cancellation provision, and total commitments exceed $22B.
- Execution has been strong and consistent, with 8 straight quarterly EPS beats from +2.3% to +32.8%, including $25.11 versus $20.98 in June 2026.
- The balance sheet is unusually clean for a capital-intensive memory cycle, with Debt/Equity 0.06, Debt/EBITDA 0.34, Current ratio 3.42, and $17.6B FCF.
- Capital intensity is surging. FY2026 CapEx is being raised to about $27B, versus deep-dive capex of $7.8B, and FY2027 CapEx will run above the low-to-mid 40% of revenue range.
- Startup costs will hit near-term profitability, with $100M to $200M per quarter of greenfield costs in FY2027 and a 0.5 to 1+ point gross margin drag.
- The stock has already rerated sharply, trading 63.7% above its 200-day average and only 13.8% below its all-time and 52-week high of $1255.00, leaving less room for disappointment into earnings in 4 days.
- Valuation is mixed. MU screens cheap on earnings at 24.11x P/E, but expensive on sales at 13.5x P/S versus a 9.0x peer average, which matters if margins flatten.
| Metric | Value | Context |
|---|---|---|
| Price target | $1379.95 | Ground-truth base case, 27.5% vs current $1082.28, Low confidence |
| Quant baseline call | BUY | Low conviction, quality score 78/100, risk score 20/100 (Low) |
| FY2025 revenue | $37.4B | +46.6% YoY, strong AI memory and SSD demand |
| FY2025 EPS | $7.65 | Reported change -397.9% YoY, but recent quarterly run-rate has accelerated sharply |
| FY2025 operating margin | 19.7% | Up from 5.2% in the prior fiscal year |
| Valuation | 24.11x P/E, 13.5x P/S, 12.1x P/B | P/E below peers at 33.6x, P/S above peers at 9.0x |
| Balance sheet | Debt/Equity 0.06, Debt/EBITDA 0.34, Current ratio 3.42 | Very low leverage for a heavy CapEx cycle |
| Free cash flow | $17.6B | Supports expansion funding with minimal shareholder payout |
| Capital return / deployment | $0.0B buybacks, $0.2B dividends, $7.8B capex | Historical cash deployment, now shifting harder toward capacity build |
| Earnings track record | 8 straight EPS beats | Latest two beats were +32.8% and +19.7% |
| Price setup | -13.8% vs $1255.00 high, +63.7% vs 200-day average | Strong trend, but expectations are elevated into earnings |
The biggest risk is that FY2027 margin expansion stalls as Micron spends through a much heavier build cycle, with $100M to $200M quarterly startup costs and 0.5 to 1+ points of gross margin pressure before greenfield bit output arrives around 2028.
Fiscal Q4 earnings on September 30, 2026, with the key watch items being FY2027 CapEx, margin outlook through startup costs, and any update on HBM allocation and SCA coverage.
Ownership is crowded but stable, with the top three institutions holding 71.9% and insiders at just 0.2%. Short interest is low at 2.46%, or 27.6M shares, down 7%, with only 0.6 days-to-cover, so this is not a squeeze story. Options positioning is supportive, with a positive GEX regime and zero-gamma at $936.74, about 13.67% below spot, which suggests flows have been dampening downside rather than amplifying it.
Call commentary points to strongest momentum in HBM, non-HBM DRAM, and enterprise SSD, including a $5B enterprise SSD quarter.
This is the economic center of the story, with HBM sold into a market management now sees above supply through 2028.
No geography split was provided, so the cleaner read is product mix: AI server memory is overwhelming client and handset as the main growth driver.
Micron signaled demand far above supply through 2028 and stepped up spending to expand capacity.
Management has earned credibility. On the 2024 Q4 call, Micron said FY2025 would deliver a "substantial revenue record with significantly improved profitability" and guided FQ1 FY2025 revenue to $8.7B ± $0.2B, gross margin to 39.5% ± 1.0%, and EPS to $1.74 ± $0.08. Since then, FY2025 revenue reached $37.4B, up 46.6%, operating margin improved to 19.7% from 5.2%, and Micron posted 8 straight EPS beats, including $25.11 versus $20.98 in June 2026. The strategic direction also held: management said HBM was sold out for 2024 and 2025 two years ago, and today demand is still above supply into 2028, which suggests the AI memory call was directionally right. The one clear change is capital intensity, with FY2025 CapEx once framed in the mid-30s percent of revenue and now FY2026 CapEx lifted to about $27B with FY2027 above the low-to-mid 40% of revenue, but that reflects demand strength more than drift.
We expect cash flow growth will increase in the fourth quarter.
Our demand for HBM is well above our ability to supply not just in 2027 but even 2028, across all HBM flavors. The same is also true for non-HBM DRAM.
We had said previously it would cross $100 billion in 2028 — we now see the HBM TAM easily crossing $100 billion in 2027.
These Strategic Customer Agreements cannot be canceled. There is no provision in this agreement that allows a customer to walk away.
Between adding construction and tool installs in Q4, we're increasing our fiscal 2026 CapEx to around $27 billion. We're also going to increase CapEx substantially next year, with more than half of that increase being construction.
Micron's moat sits in scale manufacturing, advanced DRAM and HBM process know-how, customer qualification, and the simple fact that very few companies can supply leading-edge memory at volume. That moat has strengthened because AI memory is not a commodity in the old sense: qualification cycles are longer, supply is tighter, and 16 non-cancelable SCAs create a stickier demand base. Durability is good over the next 3 to 5 years, though not absolute, because memory still remains a capital-heavy industry where process leadership and supply discipline decide returns.
Capital allocation is aggressive but rational for the current shortage. Micron generated $17.6B FCF, paid only $0.2B in dividends, did $0.0B in buybacks, and is redirecting cash into supply expansion with FY2026 CapEx now around $27B. That is disciplined in the sense that it prioritizes scarce high-return AI memory capacity over financial engineering, but it also raises execution stakes because more than half of the FY2027 increase is for construction before meaningful bit output arrives in 2028.
MU is a BUY. The stock is no bargain after its run, but the business is still moving faster than the share price implies if AI memory demand remains above supply into 2028 and the SCA framework holds. The clean balance sheet, strong beat history, and locked-in customer commitments support staying long through the cycle rather than trading it like old Micron. What would change the view is evidence that startup costs are overwhelming pricing power, or that HBM demand and SCA coverage are weakening rather than extending.
≣ Full research note tap to expand
Executive Summary
BUY, Medium conviction. MU rates a BUY at Medium conviction. The stock is up hard and sits just 13.8% below its 52-week high, but the core fact still matters more: management says HBM and non-HBM DRAM demand remain above supply into 2028, backed by 16 take-or-pay SCAs and $22B+ of customer commitments. The setup is not cheap on sales at 13.5x P/S, yet it remains reasonable on earnings at 24.11x P/E versus a 33.6x peer average if FY2026 to FY2027 earnings power keeps compounding.
Investment Thesis
Micron is no longer just a cyclical memory name waiting for the next pricing turn. It is one of the few scaled suppliers at the center of the AI memory bottleneck, and management is backing that statement with numbers that matter. The strongest proof is not tone from the call. It is 16 strategic customer agreements, more than $22B of commitments, nearly $18B of cash deposits, and terms that are explicitly take-or-pay with no cancellation provision. That structure is unusual for memory and meaningfully changes revenue visibility.
The market still does not seem to fully price how different forward earnings power could look if this demand picture persists through 2027 and 2028. FY2025 EPS was $7.65, but the quarterly cadence has exploded higher with $3.03, $4.78, $12.2, and $25.11 over the last four reported quarters. Even allowing for noise in quarterly memory economics, that is the signature of a business moving into a new earnings band. On that basis, 24.11x P/E is not demanding versus a 33.6x peer average. The stock is expensive on 13.5x P/S, but that matters less if margins and mix keep shifting toward HBM and data center SSD.
The case depends on three things. First, AI memory tightness must remain real, and management is saying it does through 2028. Second, Micron must execute the capacity build without destroying returns. FY2026 CapEx is moving to roughly $27B, FY2027 goes higher, and startup costs will ding margins by 0.5 to 1+ points with $100M to $200M quarterly cost burden. Third, customer concentration and hyperscaler urgency have to keep supporting disciplined pricing rather than encouraging oversupply.
What keeps conviction at Medium instead of High is the stock, not the business. MU has rallied 63.7% above its 200-day average and sits near its highs, so near-term expectations are elevated. Still, the underlying business looks headed toward a larger, less purely cyclical, AI-linked earnings stream than the old Micron playbook assumed. That remains worth owning.
Key Metrics
| Metric | Value | Context |
|---|---|---|
| Price target | $1379.95 | Ground-truth base case, 27.5% vs current $1082.28, Low confidence |
| Quant baseline call | BUY | Low conviction, quality score 78/100, risk score 20/100 (Low) |
| FY2025 revenue | $37.4B | +46.6% YoY, strong AI memory and SSD demand |
| FY2025 EPS | $7.65 | Reported change -397.9% YoY, but recent quarterly run-rate has accelerated sharply |
| FY2025 operating margin | 19.7% | Up from 5.2% in the prior fiscal year |
| Valuation | 24.11x P/E, 13.5x P/S, 12.1x P/B | P/E below peers at 33.6x, P/S above peers at 9.0x |
| Balance sheet | Debt/Equity 0.06, Debt/EBITDA 0.34, Current ratio 3.42 | Very low leverage for a heavy CapEx cycle |
| Free cash flow | $17.6B | Supports expansion funding with minimal shareholder payout |
| Capital return / deployment | $0.0B buybacks, $0.2B dividends, $7.8B capex | Historical cash deployment, now shifting harder toward capacity build |
| Earnings track record | 8 straight EPS beats | Latest two beats were +32.8% and +19.7% |
| Price setup | -13.8% vs $1255.00 high, +63.7% vs 200-day average | Strong trend, but expectations are elevated into earnings |
Financial Strength
Micron's financial position is a real asset in this cycle. Leverage is low, liquidity looks solid through a 3.42 current ratio, and $17.6B of free cash flow gives it room to fund an unusually large expansion plan without leaning heavily on the balance sheet. Margin trajectory has already improved materially, with operating margin at 19.7% versus 5.2% a year earlier, but the next leg will be noisier because greenfield startup costs arrive before the revenue benefit. This balance sheet can absorb that bridge.
Competitive Position
Micron's moat sits in scale manufacturing, advanced DRAM and HBM process know-how, customer qualification, and the simple fact that very few companies can supply leading-edge memory at volume. That moat has strengthened because AI memory is not a commodity in the old sense: qualification cycles are longer, supply is tighter, and 16 non-cancelable SCAs create a stickier demand base. Durability is good over the next 3 to 5 years, though not absolute, because memory still remains a capital-heavy industry where process leadership and supply discipline decide returns.
Management & Guidance
Current guidance and call commentary point to stronger cash generation near term and much heavier investment after that. CFO Mark Murphy said, "We expect cash flow growth will increase in the fourth quarter." Management also raised FY2026 CapEx to around $27B and signaled FY2027 CapEx above the low-to-mid 40% of revenue range, with startup costs of $100M to $200M per quarter and 0.5 to 1+ points of gross margin pressure. Given the company's 8-quarter beat streak and the fact that the 2024 Q4 promises on FY2025 revenue and profitability were broadly delivered, this guidance deserves respect, though not blind trust because the spend ramp is large.
Segment Analysis
- Product revenue by segment (Not disclosed in the provided data): Call commentary points to strongest momentum in HBM, non-HBM DRAM, and enterprise SSD, including a $5B enterprise SSD quarter.
- Data center and AI memory (Not disclosed in the provided data): This is the economic center of the story, with HBM sold into a market management now sees above supply through 2028.
- Geographic revenue (Not disclosed in the provided data): No geography split was provided, so the cleaner read is product mix: AI server memory is overwhelming client and handset as the main growth driver.
Capital Allocation
Capital allocation is aggressive but rational for the current shortage. Micron generated $17.6B FCF, paid only $0.2B in dividends, did $0.0B in buybacks, and is redirecting cash into supply expansion with FY2026 CapEx now around $27B. That is disciplined in the sense that it prioritizes scarce high-return AI memory capacity over financial engineering, but it also raises execution stakes because more than half of the FY2027 increase is for construction before meaningful bit output arrives in 2028.
Management Execution & Track Record
Management has earned credibility. On the 2024 Q4 call, Micron said FY2025 would deliver a "substantial revenue record with significantly improved profitability" and guided FQ1 FY2025 revenue to $8.7B ± $0.2B, gross margin to 39.5% ± 1.0%, and EPS to $1.74 ± $0.08. Since then, FY2025 revenue reached $37.4B, up 46.6%, operating margin improved to 19.7% from 5.2%, and Micron posted 8 straight EPS beats, including $25.11 versus $20.98 in June 2026. The strategic direction also held: management said HBM was sold out for 2024 and 2025 two years ago, and today demand is still above supply into 2028, which suggests the AI memory call was directionally right. The one clear change is capital intensity, with FY2025 CapEx once framed in the mid-30s percent of revenue and now FY2026 CapEx lifted to about $27B with FY2027 above the low-to-mid 40% of revenue, but that reflects demand strength more than drift.
Positioning & Flows
Ownership is crowded but stable, with the top three institutions holding 71.9% and insiders at just 0.2%. Short interest is low at 2.46%, or 27.6M shares, down 7%, with only 0.6 days-to-cover, so this is not a squeeze story. Options positioning is supportive, with a positive GEX regime and zero-gamma at $936.74, about 13.67% below spot, which suggests flows have been dampening downside rather than amplifying it.
Bull Case
- AI demand is not theoretical anymore. Management said, "Our demand for HBM is well above our ability to supply not just in 2027 but even 2028, across all HBM flavors. The same is also true for non-HBM DRAM."
- Customer commitment quality improved materially. 16 signed SCAs are take-or-pay, generally 5-year, with quarterly pricing bands and no cancellation provision, and total commitments exceed $22B.
- Execution has been strong and consistent, with 8 straight quarterly EPS beats from +2.3% to +32.8%, including $25.11 versus $20.98 in June 2026.
- The balance sheet is unusually clean for a capital-intensive memory cycle, with Debt/Equity 0.06, Debt/EBITDA 0.34, Current ratio 3.42, and $17.6B FCF.
Bear Case
- Capital intensity is surging. FY2026 CapEx is being raised to about $27B, versus deep-dive capex of $7.8B, and FY2027 CapEx will run above the low-to-mid 40% of revenue range.
- Startup costs will hit near-term profitability, with $100M to $200M per quarter of greenfield costs in FY2027 and a 0.5 to 1+ point gross margin drag.
- The stock has already rerated sharply, trading 63.7% above its 200-day average and only 13.8% below its all-time and 52-week high of $1255.00, leaving less room for disappointment into earnings in 4 days.
- Valuation is mixed. MU screens cheap on earnings at 24.11x P/E, but expensive on sales at 13.5x P/S versus a 9.0x peer average, which matters if margins flatten.
Valuation & Price Target
Engine price target $1379.95 (+27.5% vs current), Low confidence.
- P/E Multiple: $1987.41 (28% weight)
- PEG (growth-adjusted): $4176.29 (4% weight)
- P/S Multiple: $374.24 (20% weight)
- Quality-adjusted: $1324.71 (20% weight)
- DCF: $417.15 (6% weight)
The rating aligns with the quant baseline BUY (Low conviction), but the conviction here is Medium because the fundamental evidence has improved faster than a blended model can fully capture. The engine's $1379.95 target is reasonable as a base case, especially given the wide spread between the high-end P/E Multiple $1987.41 and low-end DCF $417.15 outputs. The main reason not to push to High conviction is valuation and setup: MU is only 13.8% off its high, trades at 13.5x P/S versus 9.0x peers, and is heading into earnings in 4 days after a 16.0% one-month move. Analyst consensus is strongly positive, but that is context, not the basis of the call.
Risk Assessment
Risk score 20/100 (Low). The biggest risk is that FY2027 margin expansion stalls as Micron spends through a much heavier build cycle, with $100M to $200M quarterly startup costs and 0.5 to 1+ points of gross margin pressure before greenfield bit output arrives around 2028.
The Bottom Line
MU is a BUY. The stock is no bargain after its run, but the business is still moving faster than the share price implies if AI memory demand remains above supply into 2028 and the SCA framework holds. The clean balance sheet, strong beat history, and locked-in customer commitments support staying long through the cycle rather than trading it like old Micron. What would change the view is evidence that startup costs are overwhelming pricing power, or that HBM demand and SCA coverage are weakening rather than extending.
Go deeper in the app
Live charts, options analytics, the full earnings-call breakdown, price alerts, and every other stock — updated in real time.
Get Tickr