They doubled orders and raised margin guidance while the street is still debating if the cycle has started.
Thesis: BESI is moving from a cyclical equipment provider to a structural bottleneck owner in the AI packaging stack. The margin expansion to 66% proves pricing power in a capacity-constrained market. All three major memory players are now locked into the hybrid bonding adoption curve for HBM4E. This is an alpha trade on the physics of 3D stacking.
Verdict: LONG — Conviction: HIGH
Catalyst: HBM4E qualification results in H2 2026 and the 30-40% sequential revenue growth guided for Q2.
Key Risk: TSMC CoWoS allocation is the ultimate ceiling. If the foundry cannot expand capacity, BESI bonders will sit in the backlog regardless of demand.
The Tell: Richard Blickman confirmed that all three memory customers are evaluating bonders for 'one specific end customer' that 'the whole world knows.' This admits BESI is the un-hedged proxy for NVIDIA's next-gen HBM architecture. If that one customer shifts timing, the entire BESI backlog is at risk.
Friction Level: MODERATE_FRICTION — The primary dispute is whether hybrid bonding growth is additive or if TCB cannibalization in memory stacking creates a volume headwind. There is also friction regarding the 99.9% yield threshold required for HBM4E volume production.
Report not found
The report data is no longer available. Please return to the archive.