They raised the revenue guide $25M, lowered the EPS guide, and 25% of revenue sits in a jurisdiction blocking their merger.
Thesis: Demand looks real: production tool of record at a Tier 1 HBM maker, all three Tier 1 logic customers engaged on NSA, $200M in Advanced Packaging orders, and $51M OCF against $22M net income. The execution is clean. The problem is that 25% of revenue is China, rising sequentially, while the Axcelis merger waits on China antitrust approval. That is a binary you cannot size. The $200M order ships in 2027 and inflates backlog optics today without contributing revenue, and inventory built $10M while nobody confirmed cancellability.
Verdict: AVOID — Conviction: MEDIUM
Catalyst: Q3 gross margin landing in the 41-42% range despite the 75bps investment drag would force a re-rate of the 2027 margin trajectory. China antitrust clearance for Axcelis would remove the binary overhang.
Key Risk: China at 25% of revenue with Axcelis merger pending China antitrust. A denial severs both the deal and a quarter of the top line simultaneously.
The Tell: Miller volunteered that customers are sharing 'long-range forecasts out beyond 2027, some long-range forecasts that they wouldn't normally be sharing.' Unprompted over-sharing of customer forecast behavior is either genuine conviction or a tell that they need the market to believe 2027 to justify the capacity spend.
Friction Level: MODERATE_FRICTION — Bull reads the $10M OpEx and 75bps GM drag as Capital Conviction for confirmed 2027 demand. Bear reads it as Capacity Ceiling inverted: doubling output before shipments confirm the ramp. Same facts, opposite implications.
Report not found
The report data is no longer available. Please return to the archive.