Arm Shares Fall 8% as Q3 Licensing Revenue Misses by 2.9%

ARMARM

Arm’s Q3 licensing revenue rose 25% year-over-year to $505 million but fell 2.9% short of the $519.9 million analyst estimate, while total revenue reached a record $1.242 billion, beating forecasts by 1.54%. Shares plunged 8% in after-hours trading on the licensing miss and negative read-through from Qualcomm.

1. Record Quarterly Revenue Fueled by AI Demand

Arm reported record quarterly revenue of $1.242 billion for the third quarter of fiscal 2026, representing a 25.7% year-over-year increase. This performance topped consensus estimates by 1.54%, driven largely by surging demand for the company’s energy-efficient chip architectures in artificial intelligence data centers and edge computing deployments.

2. Licensing Revenue Grows but Misses Analyst Forecasts

Licensing revenue rose 25% from a year earlier to $505 million, reflecting continued adoption of Arm’s IP in next-generation processors. However, this figure came in 2.9% below the $519.9 million analysts had forecast, prompting a modest share pullback in after-hours trading despite the broader revenue beat.

3. Data Center Business ‘Exploding,’ Says CEO

Chief Executive Officer Rene Haas described the data center segment as “exploding,” noting that revenue contributions from server and AI workloads are on track to rival—and potentially surpass—those from traditional handset royalties. Haas highlighted partnerships with leading hyperscale operators and recent wins for Arm-based accelerator designs as key drivers of this rapid expansion.

4. Guidance and Industry Headwinds

On the earnings call, Arm forecast fourth-quarter revenue slightly above street expectations, citing robust pipeline activity in both cloud and edge markets. The company reiterated that smartphone royalty growth could decelerate next year amid ongoing global memory supply constraints, which have prompted handset makers to trim production plans. Management emphasized its diversified business model and continued investment in AI architectures to mitigate these risks.

Sources

FYBS2
+11 more