Designs and manufactures DRAM and NAND memory. Owns the fabs — capital intensity is the model. HBM for AI stacks is the growth engine; commodity DRAM and NAND remain the cyclical drag.
Mid-upcycle on HBM and AI memory. Trailing multiples look reasonable at 19.7× only because TTM earnings already bake in the boom; after +677% the shares sit at the 91st percentile of Micron's own five-year EV/Sales history. Bullish on the cycle into the Street's FY27 ~$249B consensus — and below the Street because memory mean-reverts and 28% capex intensity eats free cash. Working capital and capacity overbuild break first.
Micron designs and manufactures DRAM and NAND — memory chips that sit next to every accelerator and inside every phone, PC and server. It owns the fabs, the opposite of NVIDIA's fabless model, which is why capital spending runs at 28.0% of revenue and $90.3B of trailing sales converts into only $15.8B of free cash.
The growth story is HBM for AI stacks; the cyclical drag is still commodity DRAM and NAND. In FY25, DRAM was $28.6B (77.1%) of product mix and NAND $8.5B. Segment reporting switched in FY25 from business units to DRAM/NAND — the chart below keeps the older BU taxonomy through FY24 so the two series are never forced onto one axis.
Through FY24 the story was Compute & Networking pulling away from Mobile, Storage and Embedded. In FY25 Micron re-cut disclosure to DRAM $28.6B and NAND $8.5B — HBM lives inside the DRAM line. Everything on the next six pages is a judgment about how long that DRAM upcycle lasts before capacity catches demand.
These are ship-to locations, not end markets. The United States grew strongly from FY22 to FY25 as AI server demand concentrated there; Taiwan and China look softer on a three-year CAGR even where dollar volumes held up.
Four-year revenue CAGR is only 7.8% because FY23 sat in the trough — then FY24–25 ripped the other way. The open question is whether free cash follows earnings, or whether 28% capex keeps converting net income into fabs.
| FY21 | FY22 | FY23 | FY24 | FY25 | What it says | |
|---|---|---|---|---|---|---|
| Revenue | 27.7 | 30.8 | 15.5 | 25.1 | 37.4 | Cycle trough in FY23, then a two-year climb |
| Revenue growth | +29.3% | +11.0% | −49.5% | +61.6% | +48.9% | FY23 was the memory winter, not a new ceiling |
| Gross profit | 10.4 | 13.9 | −1.4 | 5.6 | 14.9 | Negative gross profit in the trough — then a snapback |
| Gross margin | 37.6% | 45.2% | −9.1% | 22.4% | 39.8% | FY25 still mid-cycle; TTM is already 72.6% |
| Operating income | 6.3 | 9.7 | −5.7 | 1.3 | 9.9 | Back above the FY21 peak in dollars |
| Operating margin | 22.7% | 31.5% | −37.0% | 5.2% | 26.4% | TTM op margin is 65.8% — the boom is in the last three quarters |
| Net income | 5.9 | 8.7 | −5.8 | 0.8 | 8.5 | Near a full recovery of the FY22 peak |
| Diluted EPS | $5.14 | $7.74 | −$5.34 | $0.70 | $7.59 | TTM EPS already $44.18 on the HBM print |
| Free cash flow | 2.4 | 3.1 | −6.1 | 0.1 | 1.7 | Thin through the cycle — fabs eat the cash |
Eight beats from eight, but wide — +2.3% to +32.8%, averaging roughly +14%. Memory ASPs move faster than the Street recalibrates; the printed surprise is large and the guide still matters more.
Against nine peers across accelerators, foundry and memory, Micron prints best-in-set operating margin (65.8%) on TTM and still trades below the peer median EV/Sales near 14.9×. Cycle skepticism is in the multiple.
| Company | Market cap | Rev growth | Gross margin | Op margin | ROIC | EV / Sales | P / E |
|---|---|---|---|---|---|---|---|
| NVIDIA NVDA | $5.39T | +65.5% | 74.1% | 64.0% | 63.0% | 21.2× | 33.9× |
| TSMC TSM | $2.17T | +33.0% | 64.2% | 56.0% | 27.1% | 13.4× | 27.5× |
| Broadcom AVGO | $2.02T | +23.9% | 67.0% | 43.7% | 19.5% | 27.4× | 68.6× |
| Micron MU | $982B | +48.9% | 72.6% | 65.8% | 44.0% | 10.7× | 19.7× |
| AMD AMD | $784B | +34.3% | 53.2% | 15.7% | 7.6% | 19.0× | 122.0× |
| Intel INTC | $510B | −0.5% | 38.9% | 0.1% | 0.0% | 9.6× | n/m |
| Seagate STX | $180B | +34.1% | 45.6% | 33.6% | 51.4% | 14.9× | 55.1× |
| Western Digital WDC | $149B | +35.7% | 48.9% | 34.9% | 40.2% | 11.5× | 15.9× |
| Marvell MRVL | $190B | +42.1% | 50.6% | 16.2% | 5.0% | 22.0× | 74.0× |
| Qualcomm QCOM | $175B | +13.7% | 54.2% | 23.2% | 17.4% | 4.2× | 19.0× |
| Peer median excl. MU | $510B | +34.1% | 53.2% | 33.6% | 19.5% | 14.9× | 44.5× |
Best margins in the set, not the richest sales multiple. NVIDIA grows faster near 21× sales; Micron prints a higher TTM operating margin near 11× sales. That gap prices cycle mean-reversion — which is why our base sits below the Street even with a bullish rating.
Three questions in order: cheap against its own history, cheap against its industry, and what would have to be true for today's price to be right. For Micron the answers conflict — cheap on trailing earnings versus peers, rich on sales versus its own five years.
Cheap on earnings, expensive on sales — against itself. EV/Sales sits at the 91st percentile of five years of daily observations (5-yr median 3.4×; today 10.7×). P/E is only the 46th percentile, near its own median, because TTM EPS already includes the boom. Versus peers the trailing P/E still looks light; the history check is the one that bites.
Accepting ~$249B in FY27 implies the HBM cycle lasts through next year. Our base keeps that revenue print and fades operating margin toward mid-cycle by FY28 — which is why $1,200 sits 23.9% below the Street mean target even with a bullish rating.
The mean target of $1,575.91 has more than doubled from a year ago as the Street marked Micron up with HBM. Our $1,200 asks for less — roughly 8× on FY27E EPS near $155, with cyclical multiple compression versus the accelerator complex.
Every assumption behind the three targets on page one is on this page. Change one and the number moves — that is the point. A target you cannot take apart is a guess with a decimal place.
| Driver | Bear | Base | Bull | Why it moves |
|---|---|---|---|---|
| FY27E revenue | $160B | $249B | $284B | Street lo / mean / hi for FY27 |
| FY27E operating margin | 25% | 58% | 68% | HBM mix vs commodity ASP collapse |
| FY27E EPS | ~$52 | ~$155 | ~$217 | Aligned to Street EPS range |
| Diluted shares, B | 1.13 | 1.13 | 1.12 | ~1.13B shares outstanding |
| Exit P / E on FY27E | 8× | 8× | 8–9× | Cyclical compression vs NVDA's 23× |
| 12-month target | $420 | $1,200 | $1,800 | −51.7% / +38.0% / +107.0% |
| FY27E EPS | 6× | 8× | 10× | 12× | 14× |
|---|---|---|---|---|---|
| $90 | $540 | $720 | $900 | $1080 | $1260 |
| $120 | $720 | $960 | $1200 | $1440 | $1680 |
| $155 | $930 | $1240 | $1550 | $1860 | $2170 |
| $185 | $1110 | $1480 | $1850 | $2220 | $2590 |
| $220 | $1320 | $1760 | $2200 | $2640 | $3080 |
The cell at $155 × 8× ≈ $1,240 sits near the $1,200 base. On an EV check, $1,200 × 1.129B shares less ~$19.7B net cash is roughly $1.34T EV / FY27E sales ~5.4× — mid-cycle for a memory name, not peak AI hardware.
Seven events over the next eighteen months, each scored for how much of the thesis it settles. The two heaviest dates are a September print that closes FY26 against $129B, and a January capex season Micron does not control.
Memory is a derived demand on the same hyperscaler cheque that settles NVIDIA. Aggregate 2027 AI server and HBM dollar growth versus 2026 decides more of this thesis than any Micron guide.
| Aggregate 2027 AI / HBM | Reads as | Target |
|---|---|---|
| HBM dollars up hard | Cycle extends; Street FY27 ~$249B looks light | $1,800 |
| Growth, not a pause | Digestion without a break; base intact | $1,200 |
| Capex / HBM cut hard | Oversupply path; bear becomes the base | $420 |
Six risks ranked by what they would cost rather than how likely they are, each with the observable that tells you it is happening. A risk you cannot monitor is not a risk, it is an anxiety.
| Insider | Role | Filings | Value |
|---|---|---|---|
| Sanjay Mehrotra | CEO | 252 | $227M |
| April Arnzen | EVP People | 44 | $72M |
| Mark Murphy | CFO | 27 | $43M |
| Sumit Sadana | EVP CBO | 24 | $38M |
| All insiders | 19 people | 490 | $463M |
Read this as texture, not signal. Disposals run through plans after a +677% year, and $463M is five basis points of a $982B company. The strip shows persistent selling into strength — consistent with a cycle peak, not proof of one.
Bullish, but cycle-aware. A base below the Street is the point — not a soft rating.
Each case sets revenue, margin and multiple independently, then values primarily on FY27E EPS at a twelve-month horizon with cyclical exit multiples (8–9×). The headline is the base case.
Prices, consensus, targets and Form 4s from Financial Modeling Prep; fundamentals from Micron SEC filings. Peer medians use the ten-name set on page 4. Data as of 7 August 2026.
Generated from structured data — it cannot interview management or price an untold event. Institutional ownership was unavailable and is omitted rather than estimated.
Every figure above comes from Micron Technology, Inc.’s own filings and live market data, assembled and written up automatically. Ticker Alpha is building this for any listed company — so the research you need on a Tuesday afternoon takes a minute, not a week.