Cantor Fitzgerald’s CJ Muse predicts Micron earnings growth through 2028

Cantor Fitzgerald semiconductor analyst CJ Muse just doubled down on Micron Technology in a way that makes most Wall Street price target revisions look timid. In a research note dated June 8, Muse raised his price target on MU from $700 to $1,500, citing an intensifying memory supply crunch that he believes will power the chipmaker’s earnings well into 2028.

The core thesis: DRAM and NAND memory markets are heading into a period of sustained undersupply, and Micron sits squarely at the center of that squeeze.

The numbers behind the bull case

Muse’s projections for the broader memory industry are staggering. He forecasts total memory industry revenue reaching $1.21 trillion in 2027, broken down into $850 billion in DRAM sales and $360 billion in NAND sales. By 2028, he sees that figure climbing to approximately $1.4 trillion.

Both DRAM and NAND markets are forecast to remain undersupplied through 2028, creating the kind of pricing environment where memory manufacturers don’t just survive, they print money. Muse describes this as a setup for “stronger-for-longer earnings,” which he believes will force consensus estimates higher as the market catches up to reality.

Micron’s upcoming fiscal Q3 earnings report, scheduled for June 24, offers the first near-term test of this thesis. Analysts currently expect revenue of around $34.8 billion and earnings per share of approximately $19.72. Muse’s bull case goes considerably further, projecting that Micron’s EPS could eventually reach $200 under tightening supply conditions.

AI demand is the engine

Every major AI model, every data center expansion, every inference workload requires enormous quantities of high-bandwidth memory. Micron manufactures the HBM (High Bandwidth Memory) chips that sit alongside GPUs in AI accelerators, and demand for these components has been outstripping supply for several quarters running.

Micron has reportedly secured multi-year agreements locking in advanced memory capacities through 2027. These contracts provide revenue visibility that most semiconductor companies can only dream about.

For semiconductor investors, the raised target suggests that the AI capex cycle has more room to run than current valuations imply. A projection of $1.4 trillion in memory revenue alone by 2028 argues the opposite of any plateau thesis.

The June 24 earnings report will be closely watched for any signals about order visibility, pricing trends, and HBM-specific revenue.

Muse’s analysis contains no connection to digital assets or blockchain infrastructure. This is purely an AI and traditional semiconductor story.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.