NEW YORK – By mid-morning on Monday, June 15, the upgrade notes were arriving faster than the market could absorb them. Wells Fargo, Morgan Stanley, Goldman Sachs, Raymond James, Daiwa, and Wolfe Research — six Wall Street firms raising price targets on Micron Technology Inc. in a compressed window, some of them within hours of each other. The memory giant’s shares were up more than 8 percent before noon. The number that had been $550 at Wolfe Research less than a week earlier was now $1,250. At Cantor Fitzgerald it was $1,500. At Daiwa, $1,600.
This is what a consensus looks like when it breaks simultaneously rather than building gradually, and the question no one in the cascade is asking is the one that actually matters: what happens on June 24 when Micron reports earnings against the bar that just got raised?
The mechanics of the upgrade wave are not complicated. UBS, which set the morning in motion, said it now expects Micron’s fiscal third-quarter revenue and earnings per share to come in materially above the company’s own guidance, citing stronger-than-anticipated memory pricing. That read prompted other desks to update their models, and the models produced new targets, and the targets arrived in a cluster. Wolfe Research’s revised framework projects DRAM prices rising 200 percent in calendar year 2026 and a further 17.5 percent in 2027, with NAND pricing surging 216 percent this year and 17 percent the following year. The firm’s 2027 estimates for Micron now call for $226.5 billion in revenue and $135 in earnings per share — numbers that would represent the most profitable run in the company’s history by a significant margin.
Micron’s shares have risen roughly 223 percent year to date, with the stock having more than tripled in 2026. Wolfe analyst Chris Caso’s new target of $1,250 implies roughly 37 percent additional upside from Thursday’s trading price near $939. Cantor Fitzgerald’s $1,500 implies more than 50 percent. Those figures are not projections of a recovery — they are projections of a stock that has already tripled continuing to rally into a June earnings report that now carries expectations nobody was modeling four weeks ago.
The structural argument behind every note is the same, and it is not wrong: Wolfe Research says demand appears likely to stay ahead of supply through at least 2027 and possibly into 2028, with cleanroom capacity constraints limiting bit shipment growth, while high-bandwidth memory pricing could continue rising as suppliers try to narrow margin gaps. That framing is consistent with what Micron’s own management confirmed on its last earnings call — that the company can still fill only half to two-thirds of what its most important customers are ordering, a figure that functions less as a complaint than as a structural explanation for how 81 percent gross margins become possible in a commodity business.
The demand fulfillment constraint is real, and it is the correct variable to anchor on. A semiconductor company that cannot satisfy its customers is a semiconductor company that does not have to negotiate pricing. The problem with the upgrade cascade is not that it gets the supply-demand analysis wrong. The problem is what a coordinated re-rating does to the expectations embedded in the stock going into a binary event.
Micron reports on June 24. The guidance it gave for fiscal Q3 — approximately $33.5 billion in revenue and roughly 81 percent gross margins — was already a historic figure. The company’s prior peak for gross margins sat in the low sixties. The jump from that ceiling to the current guide is not incremental. It represents a structural repricing of what the AI memory cycle is worth. Six Wall Street firms just priced in even more.
Morgan Stanley framed an extended global memory shortage that could last another two to three years or more, while Cantor Fitzgerald argued Micron’s AI-driven memory cycle is still in its early to middle stages. Those may both prove accurate. But the question for June 24 is not whether the cycle is durable — it is whether Micron’s actual results beat the numbers that now exist in analyst models after a single morning of upgrades. UBS has already told the market it expects guidance to be exceeded. If the report merely meets original guidance, the stock will have a problem.
There is a precedent for what happens when expectations outrun results at this kind of valuation. In early June, Broadcom’s guidance miss erased more than $1 trillion from semiconductor market capitalizations in a single session despite the company’s fundamental AI business remaining intact. Micron is not Broadcom — its exposure is more directly tied to DRAM and HBM volumes than to custom ASIC design wins — but the mechanism is identical. At sufficiently elevated multiples, the spread between what the models expect and what the company delivers can compress sharply in minutes.
Wolfe Research also noted that cleanroom space will limit bit shipment growth at least through 2027, which is precisely the constraint that makes the bull thesis defensible in theory and anxiety-inducing in practice. If capacity cannot expand quickly, every dollar of incremental AI demand flows directly to pricing and margins. But Micron has committed more than $25 billion in capital expenditure for fiscal 2026, with a new fabrication facility in New York already underway and first wafers from an Idaho expansion expected by mid-2027. The investment that cures a shortage creates the next supply cycle. That clock is running.
What the coordinated upgrade cascade has produced, aside from an 8 percent gain in a single session, is a consensus so tightly packed around a single thesis that there is almost no one left to articulate the alternative. The Chinese memory industry’s $1.86 billion supply deal signed last week was the plainest recent signal that the global chip crunch is real and ongoing. It was also, read differently, a signal that competitors are moving to address it. How quickly that movement becomes capacity is the one variable the upgrade notes do not model precisely, because it cannot be modeled precisely.
Micron is now priced for an extended supercycle by nearly every major bank covering it. The structural case is compelling. Whether June 24 is the event that validates the cascade or the event that reveals how far ahead of the data the models have run — that is the question no upgrade note answers. And that question has a scheduled date: nine days from Monday morning, when the market opened with six reasons to buy and none to pause.
