They just hiked opex by 21% to chase a market they already own, while their biggest growth engine stalled at 25% mix.
Thesis: Arm is a tax on the entire AI infrastructure buildout. The 'stall' in v9 is noise; v8 is paying the bills while CSS (double royalty) ramps. Management is spending aggressively ($590M Opex Q4) because they see the pipeline. The China fade is priced in; the Hyperscaler lock-in (AWS Graviton, Google Axion, Nvidia Grace) is structural and expanding.
Verdict: LONG — Conviction: HIGH
Catalyst: FY26 guidance revealing the true ramp of CSS royalties, confirming the 'double royalty' thesis is hitting the P&L.
Key Risk: TSMC CoWoS capacity constraints capping unit volume growth, limiting royalty upside despite rate hikes.
The Tell: Jason Child trying to spin the v9 slowdown: 'The fact that it's slowed as a percentage of total... is actually a good thing.' When a CFO tells you missing a key metric is 'good,' watch your wallet. It suggests the transition is lumpier than the linear model suggests.
Friction Level: MODERATE_FRICTION — v9 adoption pace. Bulls see the stall at 25% as noise amidst v8 strength; Bears see it as a broken thesis and a 'Say/Do Gap'.
Report not found
The report data is no longer available. Please return to the archive.