
Arm Holdings shares have risen sharply year to date and remain well above the 52-week low, despite a recent pullback. A price target of $314.30 for the next 12 months implies about 3.83% upside from $302.71, with a hold recommendation and high confidence. Revenue and earnings have strengthened, including Q4 FY2026 revenue of $1.49 billion and non-GAAP EPS of $0.60, alongside more than doubling data center royalty revenue. Arm AGI CPU demand is described as exceeding expectations, with over $2 billion in committed customer demand across FY2027 and FY2028. The bull case targets $344.21, while valuation risk is emphasized by very high trailing and forward P/E multiples.
"Arm has been one of the cleanest AI infrastructure trades of 2026. Shares are up 176.93% year to date, 40.22% in the past month, and 17.91% in the past week alone, even after a 5.76% pullback on May 27. The stock sits 29% below a 52-week high of $325 and well above the $100.02 52-week low."
"The fundamentals justify the move. Q4 FY2026 revenue hit $1.49 billion, growing 20.1% YoY and beating consensus, while non-GAAP EPS of $0.60 beat the $0.57 estimate. Data center royalty revenue more than doubled YoY. CEO Rene Haas framed the year as a turning point, citing "a third consecutive year of more than 20% revenue growth" and noting Arm AGI CPU demand "has exceeded expectations.""
"Bulls have a clean narrative. Arm AGI CPU, the company's first data center chip, already has more than $2 billion in committed customer demand across FY2027 and FY2028, with Meta as lead partner. Arm holds roughly 50% CPU compute share among top hyperscalers, with NVIDIA Vera, Google Axion, and Microsoft Cobalt all built on Arm. Management is targeting a $100 billion data center CPU opportunity by 2030."
"The bear case starts with valuation. Arm trades at a trailing P/E of 382 and a forward P/E of 147, leaving little margin for error. The consensus analyst target of $232.14 sits well below the current quote. Non-GAAP operating margin co"
Read at 24/7 Wall St.
Unable to calculate read time
Collection
[
|
...
]