LRCX · Lam Research Corporation · Stock research

Current Thesis

The late-July binary resolved, and it resolved on the fundamentals. Lam's FQ4 FY26 print (2026-07-29) delivered $6.72B in revenue, +30% YoY, with gross margin at 52.0% — a 20-year high — and diluted EPS of $1.82 against $1.68 consensus. The September quarter is guided to $8.1B ±$400M versus roughly $7.09B consensus, and the calendar-2026 wafer-fab-equipment bar went from ~$140B to the low $150B range, with management citing 8–10 new fabs coming online. The stock closed 7/29 at $252.35, gapped to $303.52 at the 7/30 open, closed that session at $297.72, and printed $301.02 on 7/31.

The complication is what the tape did while that was happening. Through 7/30 the iShares Semiconductor ETF was down 27.43% for July, on pace for its worst month since September 2001, with none of its 34 holdings green — KLA -43.59%, Marvell -45.15%, Astera Labs -48.30% — while Nvidia fell 5.04% and Broadcom 1.97%. The de-rating was specific to memory, equipment and custom silicon. And the morning after a beat-and-raise, Morgan Stanley cut its target from $404 to $367, Wells Fargo from $450 to $350, and B. Riley to $350; Jefferies raised to $335, Needham held $390 and lifted estimates. Targets that clustered at $450–$500 in June now cluster $335–$390 against a $301 print.

The leg on offer has therefore changed shape. What is bought here is earnings growth against a multiple that has already been marked down, not the re-rating that carried the name to $438.50. That is a slower and more gradeable proposition than June's, and it has an unfilled gap beneath it.

Bullish and bearish views on Lam Research Corporation

The model's bull view on Lam Research Corporation (LRCX), in brief: FQ4 FY26 (2026-07-29): revenue $6.72B, +30% YoY, a company record; diluted EPS $1.82 vs $1.68 consensus. The bear view: The equipment complex de-rated violently in July. Both cases follow in full.

Bull Case

• FQ4 FY26 (2026-07-29): revenue $6.72B, +30% YoY, a company record; diluted EPS $1.82 vs $1.68 consensus. Prior guidance was $6.6B ±$400M and $1.65 ±$0.15 — the quarter cleared the top of its own EPS band.
• Gross margin 52.0%, the highest in 20 years; operating margin 38.4% against 35.0% the prior quarter. The long-term model was raised the same day to mid-50s gross and mid-40s operating margin.
• September-quarter guide $8.1B ±$400M and EPS $2.15 ±$0.15, against consensus near $7.09B and $1.83. Management framed it as more than 20% sequential growth. FY26 in total was $23.23B of revenue and $5.82 of diluted EPS.
• CY2026 WFE raised to the low $150B range from ~$140B on the 2026-07-29 call, with 8–10 new fabs cited. Etch and deposition intensity scales with NAND layer count and DRAM/HBM migration; NAND revenue doubled sequentially in the June quarter.
• CSBG (installed-base services) $2.47B, +17% QoQ and +43% YoY. This is the revenue layer that keeps running when system order timing slips.
• FY26 revenue $23.23B, up 26% from $18.44B; diluted EPS $5.82, up 41% from $4.13. June-quarter capital returns: $246M of buyback plus $325M of dividends, against $5.6B of gross cash and $2.4B of deferred revenue.

Bear Case

• The equipment complex de-rated violently in July. SOXX -27.43% through 7/30, worst month since September 2001, zero of 34 holdings green, KLA -43.59%. Lam ran from a $438.50 52-week high to a $252.35 close on 7/29 — a drawdown of more than 40%.
• Targets were cut the morning after the beat. Morgan Stanley $404→$367, Wells Fargo $450→$350, B. Riley →$350 on 7/30. Estimates went up and targets came down; the multiple is what moved.
• Trailing valuation stays demanding. Roughly 52x trailing earnings at $301.02 (7/31) on a $376.66B market cap, against the high-teens-to-low-20s Lam carried for most of the prior decade.
• The two-session move is already large and the gap is unfilled. 7/30 opened at $303.52 with a session low of $271.50 and a high of $321.18. Nothing between $271.50 and $252.35 has been retested.
• China was 26% of June-quarter revenue (Taiwan 27%, Korea 20%). Export-control changes hit bookings well before they appear in reported revenue.
• Insider supply came into the high. CEO Timothy Archer sold 30,000 shares at $390.01 on 2026-07-02 (~$11.7M) under a 10b5-1 plan adopted 2026-02-24.

Setup & Price Structure

The June frame — a rising 20-week EMA around $330–350 — stopped describing this tape once the name traded from $438.50 down to a $252.35 close on 7/29. The structure that governs now is the earnings gap. 7/30 opened at $303.52 (+20.28% from the prior close), traded $271.50–$321.18, and closed $297.72 (+17.98%); 7/31 printed $301.02. Two shelves define the reaction: $271.50, the low of the gap session, and $252.35, the pre-print close at which the gap fills completely.

Overhead, $321.18 is the first resistance from the gap session itself, and the analyst cluster sits $335–$390 — Jefferies $335, B. Riley and Wells Fargo $350, Morgan Stanley $367, Needham $390. That band sits only modestly above spot, a far thinner cushion than June's $450–$500.

Positioning observables, stated without a verdict: retail-facing coverage on 7/30 ran to "10 Stocks Driving the Move" and "Big Stocks Moving Higher," momentum framing arriving after an 18% session rather than before it. A Micron-linked ETF lost half its value in a month while remaining up nearly 2,500% over a year (7/30 headline), which measures how much unwind the memory complex absorbed without surrendering the year. The CEO's 7/02 sale at $390.01 was scheduled under a February plan. And there is no imminent company earnings date — the next Lam print is roughly late October — so near-term catalyst pressure arrives from outside the name.

Life-cycle phase: MATURING. The narrative is well known and still working — the WFE bar went up on 2026-07-29 and Needham raised estimates on 2026-07-30 — while flow has clearly moderated, with three houses cutting targets into a beat and the group posting its worst month since 2001. It is not accelerating: participation contracted across all 34 SOXX holdings in July. It is not saturated or dead either, since the order book just re-accelerated and margins reached a 20-year high.

Catalyst Calendar (next 30 days)

• 2026-08-13 — Applied Materials FQ3 FY26 print (date confirmed 2026-07-23). AMAT guided $8.95B ±$500M and EPS $3.36 ±$0.20. This is the nearest independent test of Lam's low-$150B WFE call from a second vendor.
• ~2026-08-26 (est.) — Nvidia FQ2 FY27 print. Nvidia fell only 5.04% in July while equipment names fell 40%+; its guide determines whether AI-capex enthusiasm re-broadens down the supply chain or stays concentrated at the compute layer.
• ~2026-10-21 (est.) — Lam FQ1 FY27 print. Outside the 30-day window, and the point at which the $8.1B ±$400M guide and the low-$150B WFE bar get marked. Nothing company-specific is scheduled before it.

What Would Change Our Mind

The gap is the structure that matters. The 7/30 session opened at $303.52 and never traded below $271.50; the entire post-print re-rating sits above the $252.35 close of 7/29. A weekly close below $252 fills that gap outright and says the market declined to pay for a raised WFE outlook plus an $8.1B guide against $7.09B consensus — at which point July's de-rating reads as a repricing of the cycle's terminal multiple rather than an overshoot.

Secondary conditions, each separately observable:

• Applied Materials guiding its October quarter below consensus on 2026-08-13 would put the low-$150B 2026 WFE figure in question from a second vendor.
• Lam walking the 2026 WFE number back below $150B at the ~late-October print, or guiding December revenue below the September level.
• The theme flipping to saturated: further target cuts while estimates rise, or the $335–$390 cluster compressing toward spot instead of leading it.

Running the other way, a weekly close holding above $321.18 — the 7/30 session high — with the WFE basket participating would argue that July was a positioning event, and would put the $335–$390 cluster in play as the next reference band.

Correlation Notes

• Lam sells etch and deposition wafer-fab equipment and trades with AMAT, KLAC and ASML. It is not an analog or MCU maker. KLA fell 43.59% in July; the basket moves together on WFE headlines.
• The July divergence is the structural fact to track: Nvidia -5.04% and Broadcom -1.97% against KLA -43.59%, Marvell -45.15%, Astera Labs -48.30%. AI compute leadership held while the memory, equipment and custom-silicon second derivative was cut roughly in half. Lam sits in the half that was cut.
• Memory-maker prints and DRAM/NAND contract pricing remain the order read-through. June-quarter systems mix: foundry 44%, NAND 23%, DRAM 23%.
• Geographic concentration: Taiwan 27%, China 26%, Korea 20% of June-quarter revenue. Korea tape shocks and China export-control headlines both transmit directly into the order narrative.
• Hyperscaler results move this name on the same session. Microsoft's report was the proximate trigger for the 7/30 chip rebound — the Nasdaq-100's best day since May 2025, with SanDisk +24% — and preceded any Lam-specific data on that tape.