Micron Technology reports fiscal fourth-quarter results after the close on September 30, and almost nobody is heading into the print expecting ordinary numbers. This Micron earnings preview begins with a company that itself guided to roughly $50 billion of revenue, an 86% gross margin and $31 of non-GAAP EPS after a fiscal Q3 in which revenue reached $41.5 billion and gross margin hit a company-record 84.9%. Current consensus has moved slightly higher, to roughly $50.9 billion of revenue and around $31.4–$31.5 of EPS.
The obvious debate is therefore whether Micron can beat numbers that already look extraordinary. But that may be the wrong question.
After a more than 280% rise in 2026, investors have already spent months pricing in AI-driven memory scarcity, massive pricing power and an unusually profitable cycle. Wednesday’s real test may be whether Micron can convince investors that today’s economics survive the next two quarters—and eventually 2027. That puts one number above almost everything else: forward gross margin.
The Priced-In Scenario: Micron Earnings Preview
Wall Street enters Wednesday with an unusually aggressive base case. Micron’s own Q4 outlook calls for approximately $50 billion of revenue, plus or minus $1 billion, while non-GAAP EPS is expected around $31, plus or minus $1. More important, management guided gross margin to approximately 86%, extending the remarkable progression from the record 84.9% achieved during fiscal Q3.
Those numbers are being supported by pricing rather than merely explosive shipment growth. In Q3, DRAM revenue increased 67% sequentially even though bit shipments rose only in the low-single-digit percentage range, as prices increased in the low-60s percentage range. NAND was even more dramatic: revenue almost doubled sequentially while prices increased in the mid-80s percentage range.
That makes another giant quarter relatively easy to imagine. Management has said DRAM and NAND demand continues to significantly exceed industry supply and expects tight conditions to persist beyond calendar 2027. AI infrastructure remains the central demand engine, while HBM consumes disproportionate wafer capacity and consequently constrains conventional DRAM supply.
But this is exactly where expectations become dangerous.
Micron’s stock has risen more than 280% during 2026 even after declining more than 6% during the current quarter. The market already understands the shortage, the AI demand story and the pricing environment. The surprise threshold is no longer “strong memory pricing”; investors are already paying for extraordinary profitability to persist.
And management itself planted the first seed of caution last quarter. Its Q4 gross-margin outlook reflected what CFO Mark Murphy called a “meaningful moderation” in the rate of price increases. That does not mean margins are about to deteriorate. It means investors now need to distinguish between a slower ascent and the beginning of a peak.
That distinction sits at the heart of this Micron earnings preview because the market is no longer debating whether conditions are strong, but how long those conditions can remain exceptional.
The Metric That Actually Matters: Forward Gross Margin
Revenue and EPS will dominate the earnings headlines, but forward gross margin is the cleaner test of whether Micron’s earnings power has structurally changed.
The mechanism is straightforward. Micron generated an 84.9% consolidated gross margin in Q3 and expects approximately 86% in Q4. At these levels, additional pricing gains do not translate into the same incremental margin expansion they did earlier in the cycle. Murphy explicitly acknowledged that dynamic during the previous earnings call.
What happens next therefore matters enormously.
If Micron can guide margins above the Q4 level—or otherwise signal that margins can remain around these extraordinary levels despite moderating price increases—the market gets evidence that this isn’t simply another commodity-memory spike. That would strengthen the argument that AI, HBM mix, constrained wafer capacity and long-term customer agreements have altered the economics of the business.
The strategic customer agreements make this question particularly interesting. Micron had signed 16 SCAs by its previous earnings call, covering roughly 20% of DRAM volume and one-third of NAND volume over their terms. Fourteen represented approximately $100 billion of cumulative minimum contractual revenue, while the agreements signed so far carried approximately $22 billion of cash deposits and related financial commitments.
More importantly, management said gross margins at the contractual floor prices would remain well above Micron’s peak quarterly margins in previous cycles.
That statement gives Wednesday’s margin commentary unusual significance. Investors will want evidence that current spot-market scarcity can eventually transition into contractual profitability without producing the familiar memory-cycle collapse. For the Micron earnings preview, this is what makes forward gross margin more consequential than a modest beat or miss against headline consensus.
This also explains Micron’s peculiar valuation profile. As of September 25, the company traded around 24.4x LTM diluted EPS but just 7.4x NTM normalized earnings. LTM EV/EBITDA stood around 17.6x versus only 5.8x forward.
Those aren’t contradictory numbers. They show how violently earnings expectations have moved.
If the forward margin structure proves durable, today’s apparently low forward multiple has a fundamentally different meaning. If margins are close to cyclical peak levels, forward earnings estimates become vulnerable and that multiple can rise rapidly without the stock price moving at all.
Micron’s near-term numbers may remain exceptional, but the investment debate has shifted from magnitude to durability. Forward gross-margin guidance, customer agreements and the pace of future capacity expansion will determine whether today’s low forward multiple reflects sustainable earnings or peak-cycle expectations.
What An Upside Surprise Would Actually Look Like
The clean upside scenario isn’t Micron reporting $51 billion instead of $50 billion of revenue. Given current expectations, investors may view that simply as confirmation of what they already believe.
A more consequential surprise would be management indicating that gross-margin expansion has further room to run, accompanied by evidence that supply remains unable to catch AI-driven demand.
There are several mechanisms already supporting that possibility. Micron said last quarter that DRAM and NAND conditions should remain tight beyond calendar 2027. HBM4’s 12-high ramp was progressing twice as fast as HBM3E 12-high, and Micron had already shipped more than $1 billion of HBM4 revenue. Management also expects HBM’s increasing trade ratio to continue consuming manufacturing capacity that otherwise could have supplied conventional DRAM.
Then there are the SCAs. Customers aren’t merely telling Micron they expect demand to remain strong; some are making binding, multiyear take-or-pay commitments and supplying substantial deposits. Micron eventually expects approximately half or more of company revenue to fall under strategic customer agreements.
A stronger-than-expected forward margin outlook would connect these pieces.
It would suggest that higher-value product mix can offset moderating price increases, that supply scarcity remains sufficiently severe to protect economics and that long-term agreements are beginning to convert what historically was volatile memory profitability into something more predictable.
Psychologically, that would force investors to revisit the idea that Micron’s earnings are automatically approaching a cyclical peak simply because margins have reached unprecedented levels.
The upside surprise is durability, not another record.
That distinction matters for a stock whose forward P/E has already fallen into the high-single digits despite its enormous 2026 rally. The market doesn’t need proof that fiscal Q4 was exceptional. It needs evidence that current forward earnings estimates deserve to be capitalized rather than discounted as temporary.
That is also why the Micron earnings preview ultimately turns on what management says about the next stage of the cycle rather than how far Q4 lands above the company’s original guidance.
Where The Bull Case Could Crack
The downside scenario is almost the mirror image.
Micron could beat Q4 revenue and EPS expectations and still leave investors dissatisfied if its fiscal Q1 outlook suggests that 86% gross margin was closer to a ceiling than a waypoint.
Management has already warned that the rate of price increases should moderate. That is perfectly manageable while supply remains exceptionally tight and richer product mix offsets slower pricing. But if forward margins flatten materially or begin declining sooner than investors expect, the interpretation changes.
The market could start asking whether the extraordinary pricing gains in Q3 and Q4 pulled forward too much profitability.
That question becomes especially relevant because additional capacity is coming. Micron expects first wafer output from ID1 in Idaho around mid-calendar 2027, meaningful shipments from its Tongluo facility around the same period, and additional HBM packaging capacity from Singapore during the first half of calendar 2027. Management has also indicated that industry supply should begin improving gradually during 2028.
None of this means an immediate glut. Micron has repeatedly argued the opposite, saying it still lacks visibility into when supply can actually catch demand.
But expectations leave limited room for deterioration. Investors currently have to believe both that enormous AI infrastructure spending persists and that memory supply remains constrained enough to preserve exceptional pricing economics.
If either side weakens, earnings estimates can reset quickly.
That is where the valuation paradox becomes important. Micron’s approximately 7.4x NTM normalized P/E looks inexpensive only if the “E” remains durable. A forward multiple built on peak earnings can be far less cheap than it appears.
Wednesday’s risk, therefore, isn’t necessarily a bad quarter. It is a great quarter accompanied by the first evidence that the next one becomes harder. That is the central downside tension running through this Micron earnings preview.
Beyond Wednesday: The Next Two Calls Matter More
The most important signals after this print will revolve around whether Micron can turn a historically cyclical shortage into a structurally more durable earnings model.
Start with the SCAs. Management eventually expects half or more of company revenue to be covered by these agreements, while arrangements containing fixed prices or ceilings at or near Q2 market prices are expected to represent approximately 40% of revenue when planned agreements are completed. Investors should watch how quickly those agreements become economically meaningful and whether management continues emphasizing margins at contractual floors.
HBM is the second structural test. Micron has deliberately targeted HBM market share roughly in line with its broader DRAM share rather than maximizing HBM volume regardless of consequences. That matters because HBM consumes substantially more wafer capacity, tightening non-HBM availability while giving Micron exposure to one of AI infrastructure’s highest-value memory categories.
Then comes capacity.
Micron expects materially more volume beginning around mid-2027 and growing into 2028. Those ramps initially create startup costs but eventually provide manufacturing absorption and additional supply. Investors will need to determine whether AI demand is expanding rapidly enough to absorb that capacity without undermining pricing.
Finally, watch the demand mix itself. Management has described AI as moving beyond accelerator racks into CPU infrastructure, storage, edge devices, automotive systems and eventually robotics. If that broadening materializes, Micron becomes less dependent on any single AI spending category.
The next phase of the thesis depends on demand broadening faster than supply normalizes.
That is a six-to-12-month question, not something Wednesday’s headline EPS number can settle. The longer-term value of this Micron earnings preview therefore lies in identifying whether those structural signals continue strengthening after the immediate earnings reaction fades.
Durability, Not Another Record, Now Defines Micron’s Earnings Debate.
Conclusion: Micron Has To Beat Expectations About Expectations
Micron heads into September 30 with perhaps the enviable problem of having performed too well.
Revenue has reached successive records. Gross margin has climbed to unprecedented territory. AI demand has turned memory into a strategic constraint for customers, while long-term agreements are beginning to give Micron visibility that would have looked unusual in previous memory cycles.
But the stock market is forward-looking, and much of that transformation is already embedded in expectations.
The contrast between roughly 24.4x LTM earnings and 7.4x NTM normalized earnings captures Wednesday’s entire debate. Investors aren’t primarily deciding what Micron earned. They are deciding how much confidence to place in the extraordinary earnings currently projected ahead.
That makes interpretation more important than the headline beat.
A quarter above consensus accompanied by resilient or expanding forward gross margins could reinforce the argument that AI has fundamentally altered Micron’s earnings structure. A similarly spectacular quarter paired with softer margin signals could revive the old memory-cycle question surprisingly quickly.
Wednesday’s most important number may therefore be the one Micron gives investors for the quarter that hasn’t happened yet.
Disclaimer: We do not hold any positions in the above stock(s). Read our full disclaimer here.
