Data center royalties doubled, yet management is aggressively pivoting to low-margin design services to pad the top line.
Thesis: ARM is trading its pure, high-margin IP leverage for a capital-intensive product fight. The 21% royalty growth is legitimate, but the $178M/quarter SoftBank injection effectively subsidizes the optical growth rate. When that 'funded R&D' turns into actual silicon, ARM competes with its own customers. You don't pay 100x earnings for a design services shop masking itself as a monopoly.
Verdict: AVOID — Conviction: HIGH
Catalyst: The inevitable transition of SoftBank 'license' revenue into lower-margin product revenue in ~12 months.
Key Risk: Hyperscaler capex exhaustion. ARM's growth is entirely derivative of NVDA/Google/MSFT spend. If they sneeze, ARM catches pneumonia.
The Tell: CFO Jason Child admitting the SoftBank revenue is 'cannibalistic' to current streams. He explicitly framed the $178M not as organic demand, but as 'funded R&D' that disappears when the product launches.
Friction Level: HIGH_FRICTION — The quality of revenue. Bulls see total growth; Bears see 15% of revenue as a related-party R&D subsidy that cannibalizes future margins.
Report not found
The report data is no longer available. Please return to the archive.