CEVA Q3 revenue misses estimates; sees strong Q4

Overview

• Silicon IP licensor's Q3 revenue rose 2% yr/yr but missed analyst expectations

• Company achieved record AI licensing revenues, over 35% of licensing and related revenues

• Royalty revenues grew 3% yr/yr and 18% sequentially, driven by modem ramp at U.S. customer

Outlook

• Company expects a strong Q4 2026 for both revenue and profit

• Company sees multi-year smartphone royalty growth opportunity through 2027 and beyond

• Company expects to begin earning initial NPU royalties as AI business expands

Result Drivers

• AI LICENSING - Record AI licensing revenues contributed over 35% of licensing and related revenues

• MODEM RAMP & CONNECTIVITY - Sequential royalty revenue growth was driven by modem ramp at a major U.S. customer and strength in IoT and connectivity, partly offset by weakness in memory-related consumer devices

• NEW LICENSING AGREEMENTS - Ten IP licensing deals signed, including expansion with new and existing customers in AI and connectivity

Company press release:

Key Details

Analyst Coverage

• The current average analyst rating on the shares is "buy" and the breakdown of recommendations is 9 "strong buy" or "buy", 1 "hold" and no "sell" or "strong sell"

• The average consensus recommendation for the semiconductors peer group is "buy"

• Wall Street's median 12-month price target for Ceva, Inc. is $42.00, about 55% above its November 6 closing price of $27.10

• The stock recently traded at 37 times the next 12-month earnings vs. a P/E of 47 three months ago

Reuters Recommended Reads

• Nov 6 - TSMC considers making advanced memory chips, The Information reports

• Nov 9 - JX Advanced Metals raises full-year profit estimate on robust chip material demand

• Nov 9 - French Ardian raises over $1 billion to invest in semiconductor makers

For questions concerning the data in this report, contact Estimates.Support@lseg.com. For any other questions or feedback, contact reuters.support@thomsonreuters.com.

(This story was created using Reuters automation and AI based on LSEG and company data. It was checked and edited by a Reuters journalist prior to publication.)