#MicronReportQ4Earnings
Micron Technology (NASDAQ: MU) closed out fiscal 2026 on September 30 with the biggest quarter in its history, and the two-day reaction says more about positioning than about the numbers. Revenue landed at $54.23 billion, up 379% year over year and roughly 31% quarter over quarter, a sixth consecutive record. Non-GAAP EPS was $33.42 against roughly $31.3 to $31.6 expected, a beat of about 6%, and non-GAAP gross margin hit 87.0%, up 210 basis points sequentially from 84.9%. GAAP net income was $37.7 billion, with non-GAAP operating income of $44.64 billion, or 82.3% of revenue. My read: this is not a cyclical top being priced in. It is the market slowly accepting that memory has become a strategic AI input rather than a commodity that gets re-rated every two years.

The print mattered less than the guide, and the guide was the loudest part of the release. For fiscal Q1 2027 Micron guided revenue of $61.5 billion, plus or minus $1.5 billion, against consensus nearer $56.3 to $56.8 billion, roughly 8% above the Street, with non-GAAP EPS of $38.15 versus about $36.02 expected and gross margin of approximately 86.25%. That implies about 13% sequential revenue growth into the next quarter, and management said it expects sequential growth in every quarter of fiscal 2027. When a company this large guides 13% quarter-over-quarter growth and still says supply and demand will be much tighter in 2027 and 2028 than in 2026, you are not looking at a normal cycle.

The segment detail shows where the money comes from. Cloud Memory revenue reached $16.28 billion at an 83% gross margin, Core Data Center hit $18.0 billion at 90%, and Mobile and Client contributed $13.11 billion at 90%. Data center SSDs alone approached $10 billion, more than ten times the year-ago level and more than two-thirds of total NAND revenue. Full-year fiscal 2026 revenue was $133.19 billion, up 256%, with DRAM revenue above $100 billion for the year, gross margin of 81.1% and EPS of $75.52, up 811% year over year. Cloud and data center together now account for close to two-thirds of the business. That mix shift is the whole story: the AI buildout does not need a consumer electronics recovery to keep Micron's fabs full.

Then there is HBM, which is where the AI chip and semiconductor theme actually lives for Micron. HBM revenue grew faster than total company revenue in the quarter, and management said the vast majority of calendar 2027 HBM bit supply is already contracted at significantly higher year-over-year prices. The company has signed 26 strategic customer agreements covering about 35% of revenue through 2030, with more than 75% of 2027 output committed and roughly $150 billion in remaining performance obligations. Micron is also working with Nvidia on an industry-first custom HBM4E implementation for next-generation GPUs and NVLink Fusion platforms. As the only major US-based HBM manufacturer, Micron sits directly in the path of every accelerator shipped by Nvidia (NASDAQ: NVDA) and AMD, and its 1-gamma DRAM and G9 NAND nodes are now its largest production nodes, with next-generation nodes moving to volume production in the second half of calendar 2027.

The catch, and the reason the stock wobbled right after the release, is capital spending. Micron plans to lift fiscal 2027 capex above its earlier plan to more than $50 billion, nearly double fiscal 2026's high-twenties level, with roughly $11.5 billion in fiscal Q1 and about $25 billion in the first half alone. Most of the increase is construction aimed at cleanroom space arriving in late calendar 2028 and beyond: ID1 output is targeted for mid-2027, ID2 for late 2028, a Japan DRAM expansion for late 2028, and the first New York fab for calendar 2030. A broad employee compensation increase is also expected to nudge margins below prior analyst forecasts. I treat this as the honest cost of locking in the supercycle rather than a red flag, but it is exactly the mechanism that historically ends memory cycles, so the number to respect is capacity arriving in 2028.

Now the trading picture, because that is where the next ten days get decided. Micron closed September 29 at $1,065.08, sold off 1.72% to $1,046.79 as the capex headline hit, then buyers stepped back in and pushed it toward $1,090. Recent sessions have traded a $1,022.90 to $1,098.90 range, with an intraday print as high as $1,107.86 and a previous close of $1,097.39. Liquidity is heavy: volume has run between roughly 26.5 million and 45.7 million shares against an average daily volume of about 28 million to 34 million, with the 50-day average nearer 46.5 million. Market cap sits around $1.21 trillion to $1.24 trillion, beta over the last twelve months is about 3.31, and the stock is up roughly 270% to 277% year to date, 485% to 550% over twelve months, 193% in six months and 12.4% in the past month, against a 52-week range of about $160 to $1,255. Valuation is doing a lot of arguing for the bulls: the trailing P/E is roughly 14.3 to 14.8 and the forward P/E is near 6.2, on trailing EPS of about $74 to $75.50.

Key levels I am watching, in order. On the downside, the first shelf is $1,046 to $1,055, the post-earnings pivot where buyers defended the print. Below that comes $1,022 to $1,025, the reaction low, then the psychological $1,000 line. A deeper flush would find a well-tested band at $974 to $990, the pre-earnings consolidation that also maps to the classic pivot zone, and then the heavy support at $943 to $955, roughly 11.8% below the recent high, which screens as the strongest technical floor in the current structure. If that fails, the next measured shelf sits near $873, about 19.3% lower, and a move there would be the market voting for the oversupply thesis instead of the supercycle one. On the upside, the first test is $1,098 to $1,110; clearing it with volume opens a measured target near $1,134, then $1,200 and a retest of the 52-week high at $1,255.

The analyst map is unusually wide, and that spread is itself the trade. Post-earnings, Goldman Sachs raised its target to $1,250, Wells Fargo went to $1,220, Mizuho lifted to $1,400 and Rosenblatt jumped to $1,900 from $1,500, while Morningstar cut its fair value to $700. The street average sits near $1,496 to $1,520 across roughly 49 analysts, with a high of $2,000 to $2,200 and a low in the $361 to $852 range, and about 95% of ratings land on buy. That gap from $700 to $2,200 is not analyst sloppiness, it is genuine disagreement about how long the shortage lasts. The bull math is simple: if EPS is heading toward $150 to $160 next year, even a 10x multiple is a $1,500-plus stock, and a forward multiple near 6x is the cheapest this has been during a shortage. The bear math is equally simple: capex above $50 billion seeds 2028 supply, and that is what Morningstar's $700 fair value and Michael Burry's reported put position, struck near half the current price, are betting on.

So here is how I frame the next ten days. Base case, and I give it the highest odds, is a $1,030 to $1,130 digestion range, because the market has to absorb a $54 billion quarter, a $61.5 billion guide and a capex shock, and it rarely trends cleanly while doing that. Bull case is a reclaim of $1,110 on expanding volume, which sets up $1,150 to $1,200 if Nvidia and the broader chip tape stay bid; note that Micron's print lifted the whole sector, with the VanEck Semiconductor ETF near $608 and up about 69% this year. Bear case is a loss of $1,022 that opens $974 and then $943, and I want that move to come on above-average volume before taking it seriously, because a drift lower on light volume is noise. Two dates sit inside the window: the $0.15 quarterly dividend goes ex on October 14, a mechanical markdown, and the pending jobs report has been flagged as a market-wide swing factor while ten-year and thirty-year Treasury yields sit at their highest since 2002.

The signals I rely on, in order of weight. First, whether $1,022 to $1,046 holds on any flush, because that tells you institutions are accumulating rather than distributing, and the historical pattern here has been a beat, a dip and then a grind higher. Second, whether volume expands above $1,110, since breakouts in high-beta memory names without volume tend to fail. Third, the forward multiple near 6x, the strongest valuation argument on the board, which only works if EPS keeps compounding, so HBM4 qualification updates and any 2027 pricing news are the fundamental triggers to track. Fourth, management's own capex commentary, because another raise without matching demand signals is the one thing that would flip me from constructive to cautious on a 2028 view.

My verdict: Micron is the cleanest, highest-beta expression of two things at once, AI infrastructure demand and the onshoring of critical semiconductor supply, and a record quarter with a guide 8% above consensus and more than 75% of next year's output already sold deserves more than a two-day reaction. Fiscal 2026 revenue of $133.19 billion, up 256%, and EPS of $75.52, up 811%, are not the numbers of a company at the end of a cycle. But with a 3.31 beta, a 485%-plus twelve-month run and a legitimate 2028 supply debate, position size and stop placement matter more than conviction. Let levels, not opinions, decide your entries: $1,046 to $1,055 and $1,022 to $1,025 on the way down, $1,110 and $1,134 on the way up, with the street's $1,496 to $1,520 average as the medium-term magnet and $943 to $955 as the line that would genuinely damage the bull case.$MU