They just delivered 27% royalty growth and immediately guided it down to low teens for the next quarter.
Thesis: ARM is a cyclical royalty business masquerading as a structural AI play. The 27% growth was a comp anomaly (lapping MediaTek). Management is aggressively pivoting the narrative to 'Agentic AI' and inference to distract from the fact that training CapEx (GPUs) doesn't directly drive their royalty leverage. The CSS mix shift is real but priced for perfection.
Verdict: AVOID — Conviction: MEDIUM
Catalyst: March 24 event. Management teased 'other possible offerings' that could move numbers. If this fails to deliver tangible revenue drivers, the premium valuation collapses.
Key Risk: Smartphone unit volumes. If the 15-20% decline due to memory prices hits premium tiers (not just low-end as management claims), the '2-4% impact' model breaks.
The Tell: Rene Haas spontaneously quoting Masa ('not interested in selling one share') without being asked about specific insider sales. It reads as defensive posturing against lock-up fears rather than a natural part of the business update.
Friction Level: HIGH_FRICTION — Durability of growth. Bulls see a structural mix shift to CSS/v9 protecting margins. Bears see a one-time comp anomaly and a narrative pivot to inference to mask lack of training leverage.
Report not found
The report data is no longer available. Please return to the archive.