They beat earnings by 10% because China panic-bought legacy tools, not because the AI cycle has fully engaged.
Thesis: The street is obsessing over a temporal margin dip caused by mix normalization, missing the structural lockout. Lam has 90,000 chambers installed—a massive switching-cost moat that prints cash via spares even when tool sales lag. They are using the China windfall to fund the R&D required for the next bottleneck: backside power and gate-all-around. The downside is capped by a $1.69B deferred revenue fortress; the upside is mispriced because the market thinks they are just a memory cycler, not an AI infrastructure utility.
Verdict: LONG — Conviction: HIGH
Catalyst: Deferred revenue normalization in 2024. As the $1.69B balance bleeds down to $1B, it smooths the P&L, proving the earnings floor is higher than the street models.
Key Risk: China mix normalization happens faster than memory spending recovers, creating a 'gap year' where margins compress before volume returns.
The Tell: Doug Bettinger on R&D: 'R&D, quite frankly, has to follow a cadence independent of the level of revenue sometimes.' They are decoupling spend from near-term results. This is either supreme confidence in the tech inflection or a disregard for near-term margin protection. I read it as the former.
Friction Level: MODERATE_FRICTION — Sustainability of China demand. Bulls see structural regionalization (new fabs forever); Bears see a one-time geopolitical pull-forward that leaves an air pocket in 2024.
Report not found
The report data is no longer available. Please return to the archive.