Management promised 2025 operating leverage, then guided Q1 expenses up 20% while revenue drops 10%.
Thesis: Teradyne is swapping high-margin, stable legacy volume for lumpy, capital-intensive AI tooling. The 'record' Memory quarter is a fake-out driven by initial HBM tooling that management admits won't repeat in '25. Meanwhile, they are structurally increasing OpEx by 15-20% to serve VIPs just as their top line faces a seasonal air pocket. This is margin compression sold as investment.
Verdict: SHORT — Conviction: MEDIUM
Catalyst: January 2025 outlook update. When they formalize the 'flattening' HBM TAM and the OpEx reality hits the full-year model.
Key Risk: VIP Compute TAM upgrade in Jan could be massive (>$500M), squeezing shorts if Blackwell complexity drives higher-than-expected test intensity.
The Tell: Greg Smith explicitly capped the HBM bull run: 'The 2024 Memory TAM includes a big slug of initial tooling... It's likely that the HBM-TAM will flatten or reduce in 2025.' He killed the growth curve himself.
Friction Level: HIGH_FRICTION — Bulls see VIP Compute as a secular growth engine ($300M to $500M). Bears see a lumpy, capex-dependent niche that can't offset the exploding cost structure.
Report not found
The report data is no longer available. Please return to the archive.