AI fever keeps dragging semiconductors back into the spotlight, and Micron is the current exhibit A. Baird just raised its price target again, the stock is still carrying an absurd 2026 gain, and traders are already staring down this week’s earnings print like it owes them money.
Baird reiterated its Outperform rating and lifted its price target to $1,520 from $1,280, with analyst Tristan Gerra pointing to demand and supply trends that still favor the memory maker. The new target implies roughly 40% upside from Friday’s close, which is a polite Wall Street way of saying the stock still has room to run before reality shows up.
Trading profile: Micron was down 2.13% in pre-market trading on Sept. 29, 2026, heading into its fiscal fourth-quarter report due Wednesday, Sept. 30, 2026. That pullback does not erase the bigger picture: the stock is up 279% in 2026 and hit an intraday record high of $1,255 on June 25. Investors are clearly treating this one like both a growth story and a stress test for how much AI optimism the tape can digest.
Key takeaway: The bullish case is not built on fairy dust. Gerra cited a surge in agentic AI demand, a deceleration in industrywide DRAM supply bit growth for 2027, and a higher expected HBM margin profile next year. The earnings report will decide whether this is still a clean trade or just another expensive confession.
Micron’s note matters beyond one ticker because Gerra’s call leans hard on the memory cycle, especially DRAM supply discipline. He said industrywide DRAM supply bit growth is expected to slow in 2027 versus this year, which is Wall Street’s favorite kind of story: a future constraint that sounds precise enough to model but vague enough to survive a bad quarter.
Trading profile: There are no separate volume or price figures here for the rest of the DRAM pack, so the clean read is thematic rather than mechanical. Still, when Micron moves on memory expectations, the surrounding names tend to get dragged into the narrative like interns into a midnight deck rewrite. Investors should watch whether the market starts paying more for scarcity than for growth.
Key takeaway: If DRAM supply really tightens as projected, the group can keep acting like a royalty stream with better marketing. If that thesis cracks, the whole trade loses some of its magic and becomes what it always was: a cyclical business wearing an AI Halloween costume.
Gerra also flagged a higher expected HBM margin profile next year, which keeps high-bandwidth memory in the sweet spot of the AI buildout. HBM is where investors have decided to park their imagination, because it sits close enough to AI infrastructure to sound strategic and far enough from consumer gadgets to avoid boring people to death.
Trading profile: The evidence pack does not give separate names, prices, or volumes for other HBM players, so the smart move is to treat this as a segment call, not a stock-by-stock victory lap. But the message is clear: if margins improve next year, the market will keep rewarding anything that looks like a bottleneck inside the AI supply chain.
Key takeaway: HBM remains a favorite way for bulls to say “scarcity” without sounding like they’re just chasing momentum. The real test is whether margin expansion survives the usual chain of supply additions, customer pressure, and euphoric expectations that tend to ambush hot hardware trades.
The sharpest line in Gerra’s note was the one about the “recent surge in agentic AI demand,” which he said is notably driving CPU demand and should keep doing so next year. That matters because it ties Micron’s memory story to a broader compute buildout, the kind of setup that keeps every chip bull reaching for the same PowerPoint buzzwords.
Trading profile: The evidence pack does not identify which specific companies are capturing that demand, so this is a sector read, not a stock picker’s fantasy league. But the implication is simple: as long as agentic AI stays hot, the market will keep rewarding the picks-and-shovels names tied to servers, CPUs, memory, and the plumbing around them.
Key takeaway: This is still one of those trades where the narrative can outrun the calendar. Investors should remember that “ongoing next year” is analyst-speak for “we like the story now and we’re hoping it survives contact with reality later.”
The analyst backdrop around Micron remains aggressively friendly. LSEG consensus called for year-over-year earnings growth of 939% for the fiscal fourth quarter, and 46 of 49 analysts rate Micron a buy or strong buy, per LSEG. When nearly the whole street is leaning the same way, that is either conviction or the kind of crowded setup that makes people pretend they were cautious all along.
Trading profile: This is not a fresh print on volume, but it is still a market-moving frame. A stock up 279% in 2026 with an earnings multiple that will be judged against a 939% growth consensus is the sort of thing that can stay heroic right up until one line item disappoints and everyone becomes an expert in “expectations management.”
Key takeaway: Consensus is powerful until it is embarrassing. Micron still looks like one of the cleanest ways to express AI memory demand, but the setup is now mature enough that even good news may only buy time instead of more euphoria.
Micron sits right at the intersection of AI enthusiasm and old-school chip cyclicality, which is how you get a stock up 279% in 2026 and still trading like the market wants another lap. Baird’s higher target reinforces the bull case, but the real catalyst is Wednesday’s fiscal fourth-quarter report, because eventually every story has to answer to numbers.
For investors, the message is less “buy anything with AI in the pitch deck” and more “know which bottleneck you’re actually owning.” In this tape, memory, DRAM supply, HBM margins, and agentic AI demand are the whole sermon; the market will decide if Micron is still a cathedral or just an expensive tent.
