LRCX · Q3 2026 Earnings

Q4 FY2025 "deposition and etch criticality in the AI era" → Q1 FY2026 "AI as quantified SAM expansion with tariff/China headwinds" → Q2 FY2026 "growth exceeding our own investor-day plan" → Q3 FY2026 "multi-year outperformance setup with bias to upside."

The WFE number keeps moving up, and the framing keeps tightening. Three quarters ago Archer wouldn't put a number on CY2026. Last quarter the number arrived at $135B. This quarter it's $140B "with a bias to the upside" — a 3.7% revision in 90 days on a market that just went through a 23% step-up. The anchor sentence is the press-release qualitative: "we expect WFE of $140 billion with a bias to the upside as the industry continues to work through various constraints." The "constraints" framing matters — management is signaling that the binding limit is cleanroom and supply chain, not demand. That's the language of a setup where 2027 absorbs the spillover from a 2026 that physically cannot fit it all.

NAND went from "$40 billion over several years" to "majority before end of 2027" in two quarters. Last quarter Archer disclosed that a major customer had announced a NAND-heavy FAB and that NAND demand was growing faster than expected. This quarter the timeline collapsed: "we now anticipate that this conversion will be pulled forward, with the majority of spending occurring before the end of calendar year 2027." Pulling 12+ months of forward spend into the next 6–7 quarters is not a tone shift — it's a structural acceleration that needs to show up in deferred revenue and lead-time commentary on the June call.

Margin tone fully inverted from December. The December guide cut gross margin 210bps and operating margin 200bps, framed as tariff and mix. Three quarters later, June is guided UP 70bps on gross margin to 50.5% and UP 150bps on operating margin to 36.5%. Management's framing — "keep it roughly in the levels that we just guided you to in June. This is going to kind of level out at where it's at, I think, for the rest of the year" — repositions 50.5% gross margin from aspirational ceiling to normalized floor.

Customer-mix quality shift is the most underappreciated disclosure. Deferred revenue fell only $30M QoQ but customer down payments fell ~$300M ("the lowest level we've seen in nearly four years") while other balance components grew with the business. Archer's explanation — "the group of customers that generally provide the down payments aren't the ones that are growing the quickest" — implies revenue mix is migrating toward advanced-node customers (foundry/logic leading-edge, advanced packaging) and away from China legacy nodes. That is a quality-of-revenue upgrade hiding inside a balance-sheet line that looks flat.

Closing language is the conviction tell. "We are delivering on the tremendous opportunity in front of us with more to come" is unusually direct for a company that, three quarters ago, was hedging China and tariffs. Combined with the H2 CY2026 > H1 framing and the explicit 2027 setup, management is telling readers that the model needs to be pushed up, not just held.