A single-digit price-to-earnings ratio on a memory maker is not a bargain. It is a warning label. Micron Technology (Nasdaq: MU) closed at $1,016.59 on 4 September 2026, up 6.10% on the day, and on the company’s own guidance for the quarter that just ended it trades at roughly 8.3 times annualised earnings. That number looks like a mispricing. It is the oldest tell in cyclical investing: commodity producers are cheapest on trailing multiples at precisely the moment their margins peak. Micron’s gross margin went from 37.7% to 84.6% in twelve months. The entire investment question is not whether the company is executing, because it plainly is. It is whether an 84.6% gross margin on a commodity is a new plateau or the top of a curve.
Here is the part most coverage skips. Micron is not being re-rated upward on that margin – it is being re-rated downward, and deliberately. A year ago the market paid roughly 30 times earnings for a memory business earning 37.7% gross margins. Today it pays about 8 times for one earning 84.6%. The multiple has compressed by almost three-quarters while the earnings exploded. That is not scepticism about this quarter. That is the market explicitly refusing to capitalise these earnings as permanent, and pricing a reversion it cannot date. Anyone buying MU on the low P/E is taking the other side of a bet the market has already made, and should know that is what they are doing.
Key facts
- Micron closed at $1,016.59 on 4 September 2026, up 6.10%, and 16.2% below its 52-week closing high of $1,213.56 – stockanalysis.com daily closes, retrieved 5 September 2026
- Fiscal Q3 2026 revenue was $41.46bn, against $23.86bn the prior quarter and $9.30bn in the same quarter a year earlier – a 346% year-on-year increase – Micron Q3 FY26 results, 24 June 2026
- GAAP gross margin reached 84.6%, up from 74.4% in Q2 and 37.7% in the year-ago quarter – same filing
- GAAP net income was $28.24bn on diluted EPS of $24.67; operating cash flow was $25.39bn against $4.61bn a year earlier – same filing
- Q4 FY26 guidance is revenue of $50.0bn plus or minus $1.0bn, gross margin of approximately 86%, and diluted EPS of $30.73 plus or minus $1.00 – same filing
- Research and development was $1.32bn in Q3, just 3.2% of revenue – Micron Q3 FY26 Form 10-Q, filed 25 June 2026 (FinanceFeeds calculation)
- The company held $30.2bn in cash, marketable investments and restricted cash at quarter end, and generated $18.3bn of adjusted free cash flow in the quarter – same filing
What is actually happening, and why the margin is the whole story
Memory has always been the most brutally cyclical corner of semiconductors. DRAM is close to a pure commodity: buyers care about density, speed, power and price, and switching between suppliers is a procurement decision rather than an engineering rewrite. That is why the industry has historically earned 20% to 45% gross margins across a cycle, and why it periodically earns nothing at all.
What broke the pattern is that artificial-intelligence accelerators consume memory in quantities the industry never provisioned for. High-bandwidth memory sits physically beside the GPU die and is consumed per accelerator, not per server. Every incremental rack of AI compute therefore drags a disproportionate quantity of DRAM with it, and that demand arrived while the industry was still running capital discipline learned from the 2023 downturn. Supply could not answer. Price did the work instead.
The evidence sits in Micron’s own segment tables. In the Mobile and Client Business Unit, revenue went from $3.26bn to $11.52bn year on year while gross margin went from 24% to 87%. In Automotive and Embedded, revenue roughly quadrupled to $4.63bn and gross margin went from 26% to 79%. These are not units where anyone expected structural margin transformation. They are units that happened to be selling into a market where the marginal buyer had stopped negotiating. We have tracked DDR5 contract pricing up nearly 500% as hyperscalers pre-book 2027 supply, and the same squeeze is now forcing Nvidia to raise AI server prices by more than 15% on memory cost alone.
The company frames this as durable rather than cyclical. “Micron’s record fiscal Q3 financial results and even stronger outlook for Q4 reflect the strategic value of memory in the AI era,” said Sanjay Mehrotra, Chairman, President and Chief Executive of Micron Technology, in the results release. “We believe our multi-year Strategic Customer Agreements will significantly enhance the durability and predictability of Micron’s strong financial performance.”
The Strategic Customer Agreements are the actual bull case
That last sentence deserves more attention than it has received, because it is the only part of the Micron story that would genuinely change the shape of the cycle rather than its amplitude.
Memory has never had contracted revenue. It has had spot pricing, quarterly negotiation and long-term agreements that functioned as volume commitments with price left open. If Micron has now signed multi-year agreements that fix both, then the business has partially converted itself from a price-taker into something closer to a contract manufacturer with visibility. Micron describes the agreements as “transformational” in its own headline, which is not language a company uses about routine supply deals.
The caution is that the release does not disclose the counterparties, the tenor, the contracted volumes, or – critically – whether the agreements fix price or merely fix supply. Every element that would let an investor judge durability is absent. Until Micron quantifies what share of forward revenue is contracted and at what margin, the Strategic Customer Agreements are a stated intention rather than a demonstrated change in the business model. That distinction is the difference between the bull case and the base case below.
Peers are behaving as though the boom is real and finite at the same time. SK Hynix and Samsung are both expanding, and expansion is precisely what ends memory cycles. The industry’s record is unambiguous: every capacity response to a price spike has eventually overshot demand. The only question that has ever mattered is the lag.
The valuation maths, and the trap inside it
Take Micron’s own Q4 guidance at face value. Diluted EPS of $30.73 annualises to $122.92. Against a $1,016.59 share price that is 8.3 times earnings (FinanceFeeds calculation on company guidance). With approximately 1,129.4 million shares outstanding as of the Q3 10-Q cover date, the market capitalisation is roughly $1.15 trillion.
Now apply the cyclical test. The relevant comparison is not to software at 30 times or to Nvidia. It is to dry-bulk shipping in 2022 and to nitrogen fertiliser in the same year – businesses that printed three to five times normal earnings, traded down to four to six times those earnings, and were correctly priced the entire time. In each case the low multiple was the market’s estimate of how much of the earnings was real and recurring, and in each case the market was closer to right than the buyers of the apparent bargain.
Here is the arithmetic that sets the bear case. Hold revenue at the guided $50bn per quarter and revert gross margin to 40%, which is still above the 37.7% Micron earned as recently as the May 2025 quarter. Gross profit becomes $20bn. Subtract guided operating expenses of about $1.86bn and the operating line is roughly $18.1bn, or approximately $13.3 per share per quarter after tax on the current share count. That is around $53 annualised. On a 10.5 times multiple, appropriate for a cyclical at mid-cycle, the shares are worth about $560. Note what that scenario does not require: no demand collapse, no revenue decline, no loss of share. Price normalisation alone gets there.
| Scenario | Level | What has to be true |
|---|---|---|
| Bull | $1,500 | Strategic Customer Agreements are disclosed as fixing price as well as volume across 2027, gross margin holds above 75%, and the market accepts 12x a contracted earnings stream |
| Base | $1,150 | Margins peak in the mid-80s and drift down slowly through calendar 2027; the multiple stays in single digits and earnings do the work |
| Bear | $560 | Gross margin reverts toward 40% on unchanged revenue as competitor capacity lands; no demand shock required |
The structural tension nobody is pricing
Memory is now a strategic material, and that invites the kind of attention that changes returns. Micron’s manufacturing is distributed across the United States, Japan, Singapore and Taiwan, with a significant fabrication presence in Taichung. Any serious disruption in the Taiwan Strait is simultaneously a supply shock that would spike prices and an asset risk that would impair the company. Those two effects point in opposite directions for the share price and no model resolves them cleanly.
The second tension is customer concentration in the AI complex. Micron notes HBM4 is in high-volume shipments “for our lead customer’s platform,” with qualification samples shipped to multiple end-customers and HBM4E volume production expected in calendar 2027. A lead customer that large is a revenue pillar and a negotiating counterparty at the same time. When memory supply loosens, the pricing power in that relationship changes hands quickly, and it changes hands with the customer that accounts for the most volume.
The third is that capital is arriving. Micron spent $7.1bn on net capital expenditure in a single quarter. So is everyone else. The industry is collectively building the supply that ends the shortage, and the lead time on a memory fab means those decisions are already made and largely unstoppable. The capacity that breaks the current price is being poured now.
What happens next
Three concrete expectations, with the reasoning attached.
First, the Q4 print is not the risk event. Micron’s fiscal year ended 3 September 2026 and results are due within weeks. Guidance of $50bn and 86% gross margin was set in June, in the middle of the squeeze, and the pricing environment has not loosened since. A beat is more likely than a miss. The share price reaction will depend almost entirely on what management says about calendar 2027 supply, not on the quarter itself.
Second, the disclosure that matters is contract coverage. If Micron quantifies the Strategic Customer Agreements – share of forward revenue, tenor, and whether price is fixed – the stock re-rates on that sentence alone, in whichever direction the answer points. A company earning 86% gross margins that can demonstrate two years of contracted revenue is not an 8x business. One that cannot is.
Third, watch competitor capacity announcements rather than demand data. Demand is not the variable in this cycle; it is visibly enormous and reasonably well telegraphed by AI capital expenditure plans. Supply is the variable, and it is announced in public. The first quarter in which industry bit growth outruns demand growth is the quarter the margin story ends, and it will be signalled by capacity commitments long before it shows in pricing.
Our numbers: bull $1,500, base $1,150, bear $560, against a spot of $1,016.59. What would change our mind on the bear case is a Strategic Customer Agreement disclosure that fixes pricing into 2028. What would change our mind on the bull case is a single competitor announcing a capacity expansion that lands before mid-2027. For a broader view of how the memory complex is being repriced, our SanDisk analysis covers the NAND side of the same squeeze, and BlackRock’s trillion-dollar semiconductor position shows how institutional money is sized against it.
Frequently asked questions
Why does Micron trade at only about 8 times earnings?
Because the market does not believe the current earnings are repeatable. Micron’s gross margin rose from 37.7% to 84.6% in a year, and memory has never sustained margins at that level through a full cycle. A low multiple on peak earnings is the market’s standard way of pricing a cyclical top, and it is not the same thing as a cheap stock.
What is the single biggest risk to the Micron share price?
Competitor capacity. The bear case at $560 requires no fall in revenue at all – only a return of gross margin toward 40% on the guided $50bn quarterly revenue. Since every memory participant is expanding into this price signal, supply normalisation is the base-rate outcome, and the only genuine uncertainty is timing.
What are Micron’s Strategic Customer Agreements?
They are multi-year supply agreements that Micron describes as transformational and says will improve the durability and predictability of its financial performance. The company has not disclosed counterparties, volumes, tenor, or whether the agreements fix price as well as supply, which is why they support the bull case rather than the base case.
When does Micron next report results?
Micron’s fiscal 2026 ended on 3 September 2026, so fourth-quarter and full-year results are due within weeks of publication. Guidance calls for revenue of $50.0bn plus or minus $1.0bn, gross margin of approximately 86% and diluted EPS of $30.73 plus or minus $1.00.
How much has Micron stock risen in the past year?
The shares closed at $1,016.59 on 4 September 2026 against a 52-week low of $124.21, an increase of roughly 718%. The stock nonetheless sits 16.2% below its 52-week closing high of $1,213.56, set earlier in the summer.
This article is analysis and information, not investment advice. Scenario levels are the author’s estimates based on company filings and are not price targets or recommendations. Trading and investing carry risk, including the total loss of capital. Figures were verified against primary sources on 5 September 2026 and may have moved since.