Micron Technology, Inc. (MU): what the price assumes

In the published model solve dated 2026-Q2, anchored at $930.80, Micron Technology, Inc. (MU) is priced for +26.4% growth. boothcheck publishes no house fair value, target price, or buy/sell rating; individual model outputs and user-controlled scenarios are analytical inputs, not Boothcheck targets. Narrative composed 2026-08-07.

Generated: 2026-08-30 · Exported: 2026-08-31 · Source: https://boothcheck.com/report/MU

Headline

FieldValue
TickerMU
CompanyMicron Technology, Inc.
Sector / IndustryTechnology
Current price$930.80/sh
CompositionDRAM 76% / NAND 23% / Other (primarily NOR) 1%

What The Price Assumes (Inversion)

The assumption today's price embeds, recovered by inverting the valuation.

FieldValue
Inversion basiswhole-company
Operating margin (value-band context)22.2%
Operating margin today65.6%
Margin compression (value-band)-43.4pp
Implied growth26.4%
Multiple paid18x operating income

The operating-margin figure is value-band context at year 11: derived from the framework's value band, a separate calculation — not part of the priced-in solve.

Solve inputs: computed at a 12.5% cost of capital with 4% terminal growth over a 5-year stage.

How unusual the bet is: n/a

ReferenceValue
vs own history-0.41σ

Valuation X-Ray

The price is justified by relative-multiple and growth-DCF; asset-based/earnings-power land below the price.

How the valuation models price the stock relative to the market price. Price/FV above 1.0 means the market pays more than that lens defends (expensive); at or below 1.0 the lens can defend the price.

FamilyMedian price/FVModelsReads
Asset1.93x5expensive
Earnings3.43x5expensive
Relative1.16x2expensive
Growth0.77x3justifies

Families that justify the price: Relative, Growth Families that call it expensive: Asset, Earnings

The models below discount at their own flat-beta convention rates (cost of equity 9.3%, WACC 9.2%); the inversion above states its own rate.

Per-Model Detail (n=15)

ModelFamilyFVPrice/FVApplicableMethodology
DCF Perpetual GrowthGrowth$1524.910.61xyesFCF base $50.4B, growth 25% (input: historical growth), terminal g 4.0%, WACC 9.2%, 7yr projection
DCF Exit MultipleGrowth$1114.990.83xyesExit EV/EBITDA: 12.5x / 15.5x / 18.5x (bear / base = today's held flat / bull), 7yr
Relative ValuationRelativenoP/E 22x (static sector reference · 2026-04), scenarios: 17.6x / 22.0x / 26.4x (bear / base = reference held flat / bull), EV/EBITDA 16x
Simple DDMGrowthno
Two-Stage DDMGrowthno
Simple Excess ReturnAsset$483.101.93xyesBV/sh $89.18, ROE (TTM) 50.1%, ke 9.3%
Two-Stage Excess ReturnAsset$1372.420.68xyes5yr excess ROE then converge to ke=9.3%
Discounted Future Market CapGrowth$1209.740.77xyesRev $90.3B, growth 30% (input: historical growth; tapered), Terminal P/S: 9.3x / 11.6x / 14.0x (bear / base = today's held flat / bull, cap 12x)
Peter Lynch Fair ValueRelative$530.041.76xyesEPS $44.17, growth 2% (input: historical EPS growth), PEG=10.41 (Overvalued)
Margin TrajectoryGrowthno
Earnings Power ValueEarnings$119.067.82xyesNormalized EBIT (5y avg op income, one-time charges added back) $14.79B × (1−21%) / WACC 9.2% → EPV (no growth)
Residual IncomeAsset$782.001.19xyesBV $89.18 + 5yr PV of (ROE (TTM) 50.1% − Kₑ 9.3%) × BV; BV grows 8.8%/yr
Graham NumberAsset$297.713.13xyes√(22.5 × EPS $44.17 × BVPS $89.18) — Graham's conservative floor
EV/EBITDA RelativeRelativenoEBITDA $66.11B × sector EV/EBITDA 16.0x
FCF YieldEarnings$271.563.43xyesFCF $26172.0M / Kₑ 9.3% — zero-growth perpetuity
SBC-Adj FCF YieldEarnings$260.013.58xyesSBC-adj FCF $24.96B (FCF $26.17B − SBC $1.21B) capitalized at Kₑ
Ben Graham FormulaEarnings$1425.220.65xyesEPS $44.17 × (8.5 + 2×15.0%) × (4.4 / 5.3%)
ROIC-Justified P/BAsset$329.952.82xyesBV $89.18 × (ROIC 34.2% / WACC 9.2%)
P/Sales SectorRelativenoRevenue $90.27B × sector P/S 5.0x
PEG Fair ValueRelative$1656.380.56xyesEPS $44.17 × (PEG 1.5 × growth 25.0% (input: historical EPS growth)) → PE 37.5x
Earnings YieldEarnings$477.511.95xyesEPS $44.17 / required return 9.3% (Rf 4.3% + ERP 5.0%)
Funds From Operations MultipleRelativeno
Clinical Phase NPVGrowthno
MertonAssetno
V5 Mechanicalno

Economic-Unit Decomposition (Sum Of The Parts)

One or more material disclosed units has unresolved economics. Unknown/general is not evidence of homogeneity: segment SOTP is primary but incomplete, consolidated cash flow may remain only a secondary cross-check when every unit shares an enterprise basis, and one sector multiple or target margin is withheld.

UnitRoleValuation basisRevenueReported profitValue evidenceStatus
Cloud Memory Business Unit (CMBU)operatingenterprise$13.5b$6.1b operating-incomewithheldunresolved no unit value
Core Data Center Business Unit (CDBU)operatingenterprise$7.2b$2.2b operating-incomewithheldunresolved no unit value
Mobile and Client Business Unit (MCBU)operatingenterprise$11.9b$2.0b operating-incomewithheldunresolved no unit value
Automotive and Embedded Business Unit (AEBU)operatingenterprise$4.8b$557.0m operating-incomewithheldunresolved no unit value

No total common-equity value is stated. One or more material units lack a supported unit value. The displayed values are an indicative subtotal; consolidated debt and cash cannot be applied to a fraction of the company.

Solvency

FieldValue
Net cash$13.9b
Net debt / NOPAT (after-tax)-0.30x (net cash)
Net debt / operating income (pre-tax)-0.24x (net cash)
Interest coverage257.6x
Share count CAGR (dilution)0.5%
Burning cashno

Bullet Takeaways

Bull Case

Memory was supposed to be a commodity, and for most of its history the industry obliged: everybody built, prices collapsed, everybody stopped building. What changed is not the product. It is that the number of firms capable of making leading-edge memory shrank to a handful while the customers arrived all at once. The company describes the condition in its own filing with no adjectives at all: "The AI-driven growth in the data center has accelerated demand for memory and storage at a rate greater than our ability and the industry's ability to increase supply." Read as economics rather than as marketing, that is a shortage the seller cannot fix quickly even when it wants to.

What a shortage does to a manufacturer shows in the returns. Over the twelve months to late May 2026 the business turned 65.6% of revenue into operating profit and earned a return on equity near 50.1%. The nearest listed comparables are living the same year for the same reason: SNDK is running an operating margin near 47.4% on revenue up 106.9%, STX near 28.2% on revenue up 28.9%, and WDC near 24.4% on revenue up 32.0%. That pattern is not one management team executing unusually well. It is an entire product category repricing simultaneously, and the only firms who can sell into it are the ones whose fabs are already running.

The barrier is capital and time, and both are on the public record. Construction on a new fab in Boise began in October 2023, and the company projects "first DRAM wafer output projected in the second half of calendar 2027". Four years from ground-breaking to first wafer is the cost of entry, before a single competitive bit is sold. In June 2025 the company announced a second leading-edge fab in Idaho under amended CHIPS Act agreements. Anyone deciding today to compete for this demand arrives after the cycle that provoked the decision has already turned.

Inside the portfolio the mix is moving toward the customers who pay best. In the fiscal year ended August 2025, the cloud memory unit carried 13.5 billion dollars of revenue against 6.1 billion dollars of operating income, roughly 45 cents on the dollar. The mobile and client unit carried 11.9 billion dollars against 2.0 billion dollars, roughly 17 cents. Nearly the same revenue, about three times the profit. Every incremental wafer routed to the data center rather than to handsets moves the blended economics without any change in operating skill.

The balance sheet is being run for the downturn rather than against it. Liquid assets of 26.0 billion dollars sit against 12.1 billion dollars of borrowings, leaving 13.9 billion dollars of net cash, and operating profit covers interest about 257.6 times over. In the second quarter of this fiscal year the company prepaid five separate borrowings in full. A cyclical manufacturer retiring principal at the top of its cycle is doing the one thing that makes the bottom of the next one survivable.

Bear Case

Take the trailing numbers apart and the fragile assumption names itself. Third-quarter revenue rose 346% against the same quarter a year earlier, and almost none of that came from selling more units. The filing is exact about the split: "Sales of DRAM products increased 343%, primarily due to a low-260% range increase in average selling prices and a low-20% range increase in bit shipments." Bits went up by about a fifth. Prices went up by a multiple. Everything else in the gap is price, and price is the single variable in this industry with a documented history of travelling in both directions.

That matters because of what the quote already embeds. The market is paying roughly 17 times the operating income of the trailing year, which sounds undemanding until you ask what has to happen from there. The arithmetic requires operating profit to compound about 24.9% a year for five more years, starting from a base that is itself a record. Not holding the peak. Growing it by roughly two and a half times again. Of the companies that have ever grown at that pace, only about 34% sustained it for anything like five years, and that base rate is drawn from businesses that were not simultaneously depending on a commodity price.

There is a cleaner way to see the same problem. Averaged across the last five fiscal years with one-time charges added back, operating income comes to about 14.79 billion dollars. The trailing twelve months produced four times that. Methods that value the business on the average rather than the peak land far beneath the quote: the price sits about 3.3 times where the earnings-power methods come out and about 1.9 times where the asset-value methods do. Only the forward-growth and the peer-multiple methods reach it, and both work by treating the current year as a launch point rather than an outlier. Which of those two readings is right is the entire investment question, and the accounts cannot answer it.

Meanwhile the supply response is already contracted. Capital spending of roughly 27 billion dollars this fiscal year, net of government incentives, is capacity that will arrive regardless of whether the prices that justified it survive to meet it. The company understands the mechanism and files it as a risk: "We strive to balance our demand and supply for each technology node, but the dynamics of our markets and our customers can create periods of imbalance, which can lead us to carry elevated inventory levels and underutilized capacity." It also names the competitor that does not need to earn a return on capital, warning of "the threat of increasing competition and DRAM and NAND oversupply due to significant investment in the semiconductor industry, including by the Chinese government and various state-owned or affiliated entities, such as CXMT and YMTC".

Customer concentration compounds all of it. The company discloses that "In 2025, over half of our total revenue came from our top ten customers." and that "approximately one-half of our total revenue was concentrated in the data center end market". A short list of hyperscale buyers negotiating with a supplier currently keeping roughly 65 cents of every sales dollar have both the motive and the purchasing leverage to reset terms the moment supply loosens. The most fragile thing in this price is not demand for AI. It is the assumption that memory pricing at these levels is a plateau rather than a spike.

Valuation

Seventeen times a year's operating income, for a company whose revenue tripled, reads like a bargain. The number is real. What the multiple leaves out is where its denominator came from. The trailing twelve months to late May 2026 produced 59.2 billion dollars of operating profit on 90.3 billion dollars of revenue, a margin of 65.6%, and that margin was manufactured by a shortage rather than by a change in how the company makes things.

What the quote actually embeds is growth from that base, not maintenance of it. The market is paying for operating profit to compound about 24.9% a year across a five-year stage. Set against how often such a pace has persisted, only about 34% of comparable fast-growers kept it up for roughly five years, and the comparison set behind that figure is thin: three of the four available checks had usable data, and there is no read at all on where this multiple sits within its own peer range. That is a directional reading, not a measurement, and a missing check makes the claim weaker rather than gentler.

The methods split cleanly along the same seam. The forward-growth methods and the peer-multiple methods both reach the quote, because both take the current year as their starting point. The asset-value and earnings-power methods land well beneath it, with the price sitting about 1.9 times where the asset-value methods come out and about 3.3 times where the earnings-power methods do. The earnings-power gap is the informative one: that approach runs on a five-year average of operating income of roughly 14.79 billion dollars, a quarter of the trailing figure. Nothing about the two results is contradictory. They are answering different questions, and the spread between them is the size of the bet on which question matters.

The unit economics underneath are genuinely different from one another, which is why a single multiple across the company would mislead. In the fiscal year ended August 2025, the cloud memory unit carried 13.5 billion dollars of revenue and 6.1 billion dollars of operating income; the core data center unit 7.2 billion dollars and 2.2 billion dollars; the mobile and client unit 11.9 billion dollars and 2.0 billion dollars; the automotive and embedded unit 4.8 billion dollars and 557 million dollars. Data center memory is the profitable end and it is where the incremental capacity is being pointed. No company-wide figure is stated here because none of the individual units resolves to a defensible value on its own.

Among the peers the comparison sharpens rather than settles the point. SNDK carries an operating margin near 47.4%, STX near 28.2%, WDC near 24.4% and TXN near 37.3%, and INTC is at roughly break-even on its whole business. The three storage names are riding the same wave; the other two are not. Solvency does not bound the downside so much as extend the time available: 26.0 billion dollars of liquid assets against 12.1 billion dollars of borrowings leaves 13.9 billion dollars of net cash, and interest is covered about 257.6 times. The share count has crept up roughly 0.5% a year since mid-2022, which is compensation dilution rather than issuance. What settles this eventually is not the balance sheet or the multiple. It is what a gigabit costs two years from now.

Catalysts

The third quarter of fiscal 2026, ended May 28, 2026, was the largest step in the sequence so far. Total revenue rose 346% against the year-ago quarter and consolidated gross margin reached 85%, up from 74% in the immediately preceding quarter. Net income for the quarter was 28.2 billion dollars. Those are the figures a reader should carry forward as the peak reference point rather than as a new baseline, because the same filing attributes the move overwhelmingly to average selling prices rather than to shipments.

Spending is the counterweight and it is now committed. The company estimates capital expenditure for property, plant and equipment of roughly 27 billion dollars in fiscal 2026, net of proceeds from government incentives, and held purchase obligations of about 2.93 billion dollars for property, plant and equipment as of May 28, 2026. Longer-dated capacity is on the same track: the Boise fab that began construction in October 2023 projects first DRAM wafer output in the second half of calendar 2027, and a second Idaho fab was announced in June 2025 under amended CHIPS Act agreements.

Capital returns have stayed deliberately modest against all of this. The company declared a quarterly dividend of 0.15 dollars a share in the May quarter and declared the following quarter's on June 24, 2026. Earlier in the year it directed 650 million dollars to buying 2.5 million shares under its existing authorization while separately prepaying five borrowings in full, including the 2028 Notes and the 2030 Notes. Management is choosing balance-sheet repair over capital return while the cycle is generous, which is the behaviour of a company that expects the cycle to be a cycle.

Peer Cohorts (Per Segment, With Filing Citations)

Cloud Memory Business Unit (CMBU) / Core Data Center Business Unit (CDBU) (reported)

Mobile and Client Business Unit (MCBU) (reported)

Automotive and Embedded Business Unit (AEBU) (reported)

Methodology Note

Fundamentals sourced from SEC EDGAR filings. Current price from Databento. The priced-in inversion and valuation x-ray are computed by the boothcheck engine; narrative composed by AI from the structured data.

Sources

Q3 FY2026 Form 10-Q · Q2 FY2026 Form 10-Q · FY2025 Form 10-K

View the full interactive MU report on boothcheck