Built by Grok 4.6 High

LRCX | Earnings Preview

Reports Wednesday, October 21, 2026. Prepared Thursday, September 24, 2026. This is FY2027 Q1, the September quarter ending about September 27, and it has not been reported. Lam's IR site, checked the same day, still lists the July 29 call as the latest earnings event. FMP and the repo calendar agree on October 21. Recent calls, including July 29, were Wednesday after the close. Do not treat 5:00 p.m. Eastern as confirmed. Last reported quarter is FY2026 Q4, released after the close on July 29, 2026. Fiscal year ends the last Sunday in June. The homepage label 2026Q4 is the announcement-quarter tag and runs one quarter ahead of the quarter being reported.

FY2027 Q1 is a high bar that Lam's own history still treats as a floor, and the street has already moved onto that floor. On the July 29 call Doug Bettinger guided September non-GAAP revenue of $8.1 billion plus or minus $400 million, gross margin of 52% plus or minus 1 percentage point, operating margin of 39.5% plus or minus 1 percentage point, and diluted EPS of $2.15 plus or minus $0.15 on 1.255 billion shares. Versus the year-ago September quarter that is +52.1% revenue against $5,324 million and +70.6% EPS against $1.26. FMP on September 24 sits at $8.133 billion and $2.178, +0.4% and +1.3% above the midpoints, inside both ranges. Nothing on this page is an FY2027 Q1 actual.

Growth trajectory. The last print, FY2026 Q4, was revenue $6,722 million versus $5,171 million, +30.0% year over year, and non-GAAP EPS $1.82 versus $1.33, +36.8%. The four reported revenue growth rates into this guide were +27.7%, +22.1%, +23.8%, +30.0%. Hitting the September midpoint would reaccelerate that rate by about +2,210 basis points, and the compare is harder, not easier: September 2025 revenue already sat above June 2025. EPS growth would reaccelerate from +36.8% to +70.6%, about +3,380 basis points. The fiscal 2026 EPS rate was not a straight decline: +46.5%, +39.6%, +41.3%, +36.8%. It ended 970 basis points slower than FY2026 Q1, after a +170 basis point bounce in Q3 that Q4 gave back. Gross-margin expansion on the guide is +140 basis points versus 50.6%, narrower than June's +170 basis points. Operating-margin expansion does the opposite: +450 basis points versus +400 basis points last quarter. The bridge is opex lagging revenue, not another gross-margin step. The 52% level holds the 20-year high just printed. The expansion rate does not.

The two segments did not move together. Customer support was $2,472 million versus $1,734 million, +42.6%, and that year-ago rate has risen for four quarters (about flat, then +13.5%, +25.3%, +42.6%). Systems were $4,250 million versus $3,438 million, +23.6%, after decelerating from +48.3% in the year-ago September quarter. Inside systems, DRAM dollars roughly doubled year over year. Foundry and NAND dollars were only about +5%. Archer's "NAND more than doubled" line was versus March, not versus June 2025. Tim Archer's "more than 20% quarter on quarter" describes the guide versus June. It is his sequential framing, not the year-over-year rate.

What to watch

• Guidance, not a stale street number. Consensus has chased the July guide. Reuters on July 29 had the $8.1 billion / $2.15 guide above a roughly $7.09 billion / $1.83 consensus. WSJ market data, checked September 24, has FY2027 Q1 EPS at $2.18 now, $2.16 a month ago, and $1.80 three months ago. A print at the midpoint beats Lam's published habit on revenue by less than the street expects and disappoints the EPS habit. The December outlook issued that night, against a standing FMP bar of $8.526 billion and $2.309, is the number that is not yet in the model as a company figure.
• Margins and the tax step-up. Eight straight quarters cleared the revenue, gross-margin, operating-margin, and EPS midpoints. Three of the last four EPS prints cleared the high end of the range. This gross-margin guide starts at the 52.0% just printed, not under it. June's non-GAAP tax rate was 11.0%. Bettinger said September moves to the mid-teens on a higher U.S. mix and a higher GILTI rate, and that the uptick lasts through the rest of calendar 2026 and likely beyond. The $2.15 already absorbs that. A model that keeps 11% overstates EPS.
• NAND's year-ago rate, CSBG, and China. September is the first quarter where the NAND upgrade cycle faces an easy compare. China, the peak at $2,283 million and 43% a year ago, is the hard one. June China was $1,722 million, −6.3% year over year, with domestic customers down and multinationals in China up. Tariffs are no longer the margin item they were on the October 2025 call, when mix and tariffs took the December gross-margin midpoint down to 48.5%. They remain a live caveat, not a new dollar in this guide.
• 2027 wafer-fab equipment, still unquantified. Calendar 2026 WFE was lifted to the low-$150 billion range, stored as $150 billion, and Archer took the upside bias off. He would not number 2027. That is the gap between management's "extraordinary setup" and a street year of $34.80 billion that already assumes the September step-up sticks and then cools.

Pre-reads in front of Lam are unconfirmed: Micron on September 30, ASML on October 14, TSMC on October 15. KLA, Tokyo Electron, Teradyne, and Applied Materials report after. No close U.S. process-tool peer has a company-confirmed print before October 21. Guidance has not been withdrawn. Tim Archer and Doug Bettinger are the same pair who have been on these calls all year.

Classification: conservative versus Lam's own record. No longer aggressive versus the street. Over the last eight completed quarters the next quarter's revenue, non-GAAP gross margin, non-GAAP operating margin, and non-GAAP EPS all came in above the guided midpoint (8 for 8). None missed the low end of the range. Revenue stayed inside the high end every time. The average revenue beat of the midpoint is +2.4% over eight quarters and +2.4% over the last four. The latest, June, was the smallest of the eight at +1.9% ($6,722 million versus a $6.6 billion mid, +$122 million, inside the $400 million range). EPS beats have widened. The last four averaged about +8.7% versus the midpoint, and three of those four cleared the high end, each by $0.02 to $0.03. June non-GAAP EPS of $1.82 was $0.02 above a $1.80 high (mid $1.65). June gross margin of 52.0% cleared a 51.5% high. June operating margin of 38.4% cleared a 37.5% high (guide 36.5%).

That pattern says the midpoint is a floor Lam is willing to publish. It does not say revenue will blow through $8.5 billion. A historical revenue haircut of about +2.3% to +2.4% is a whisper near $8.29 billion, still inside the guide high and above the FMP high of $8.183 billion. A last-four EPS haircut near +9% is a whisper near $2.34, above the guide high of $2.30 and above every published FMP estimate (high $2.218). The street is not underwriting that. The entire FMP revenue band ($8.032 billion to $8.183 billion) and EPS band ($2.117 to $2.218) sits in the top half of the guide and below the guide highs. A print at the guide low would miss every published FMP estimate. A print at the guide high would beat every one. In line with the midpoint is a small beat versus history on revenue and a disappointment versus history on EPS.

The guide was aggressive versus the street on the morning of July 29. It is not aggressive now. The dollar range is the widest Lam has used, plus or minus $400 million, first used on the June guide and kept for September. As a percent of the midpoint it is about ±4.9%, similar to the old ±$300 million band when revenue was near $5 billion. The wider dollar band is scale, not a loss of precision. Margin bands stayed at ±1 percentage point.

FMP's standardized EBIT average of $3.130 billion is about 38.5% of its revenue average, roughly 100 basis points under the 39.5% operating-margin midpoint. That is a different definition, not Lam's non-GAAP operating margin. EBITDA of $3.257 billion is a street construct. Lam did not guide it. The implied non-GAAP opex of about $1,013 million ($8.1 billion times 52% minus $8.1 billion times 39.5%) is arithmetic, +21.7% versus year-ago non-GAAP opex of $832 million against revenue +52.1%. Do not treat $1,013 million as a company figure.

Daloopa stores the forward guide on the issuing quarter, fiscal 2026 Q4. It is the outlook for the unreported September quarter, not the June result. GAAP midpoints match at the rounded level: gross margin 52.0%, operating margin 39.5%, and EPS $2.15. The press release puts the reconciling items at a few million dollars, well under a cent of EPS on 1.255 billion shares. Non-GAAP is not hiding the quarter.

Lam guides one quarter, non-GAAP, and does not guide by segment. End-market figures in the history are a percent of systems revenue, not dollars. Dollar comments that use mix times systems revenue are approximate. Internal is n/a in every row. Where Lam did not guide, the percent difference versus a midpoint is n/a.

Current quarter — FY2027 Q1

The stack that matters, and that is not in the guide, is systems end-market mix above systems revenue and customer support, above consolidated revenue. Bettinger's only segment sentence was that the service profile should look like June and that everything grows. Foundry's share of systems has fallen from 60% in September 2025 to 44% in June 2026. The growth mix under a +52% consolidated guide, on the last print, was customer support and DRAM.

FQ+1 guide — FY2027 Q2, quarter ending about December 27

Not Guided. Lam has not issued a December outlook. The column is FMP for the period ending December 28, 2026. It is not a company target. Year-ago quarter is FY2026 Q2: revenue $5,345 million and non-GAAP EPS $1.27. Street $8.526 billion / $2.309 is +59.5% / +81.8% year over year, a faster rate than the September guide. The street has growth still accelerating into December. Lam has not said that. On October 21 the new guide will be compared with this street pair, not with a company number. The only color that touches the quarter after September is qualitative: 2027 drivers "largely the same as this year," ranked DRAM, then leading-edge foundry and logic, then NAND, and a mid-teens tax rate through the rest of calendar 2026.

FY+1 guide — FY2027, year ending about June 27, 2027

There is no Lam revenue, EPS, or margin guide for FY2027. Lam does not guide the year. The July 29 call did not create one. Reported FY2026, summed from the four cited quarters (5,324 + 5,345 + 5,841 + 6,722), is revenue $23,232 million and non-GAAP EPS $5.82, which matches the fiscal-year figure Bettinger cited ($23.2 billion). Do not use the stale FMP row of $23.20 billion and $5.70 on the June 28, 2026 estimate date. Street FY2027 of $34.80 billion and $9.46 is about +49.8% revenue and +62.5% EPS versus that reported year. It is consensus versus a reported year, not versus a company target. The annual low/high is wide: revenue $31.37 billion to $36.90 billion, EPS $8.56 to $10.14. The September midpoint is 23.3% of the street year. The four quarterly revenue averages sum to about $34.75 billion, in line with the annual figure. Later quarters are thin (7 analysts).

The quantified "annual" items from this call are calendar 2026 industry outlooks, updated on the call, not an FY2027 Lam P&L. Calendar 2026 WFE went from $135 billion in January to $140 billion in April to $150 billion in July, and the upside bias came off. Advanced packaging growth went from 40% to 50% to 70%. The long-term frame, not this year, is mid-50s gross margin and mid-40s operating margin over several years. September's 52% and 39.5% are still short of that destination. FY2028 ($41.68 billion / $11.86) is also Not Guided. Do not back into a 2027 WFE number from the September guide.

Eight reported quarters — periods across, metrics down

Year-over-year rates use the same fiscal quarter a year earlier. FY2027 Q1 is not on this grid.

Revenue growth troughed at +22.1% in FY2026 Q2 and reaccelerated to +30.0%. EPS growth ended FY2026 at +36.8%, 970 basis points slower than FY2026 Q1's +46.5%, on a path of +46.5%, +39.6%, +41.3%, +36.8%. Q3 bounced +170 basis points and Q4 gave it back. Tax and opex absorbed part of the revenue step. The +71% EPS guide is a break in that slowdown only if the mid-teens tax rate is already right and opex really does lag. Deferred revenue ended June at $2,430 million versus $2,681 million, −9.4%. Management highlighted a sequential rebuild. The year-ago balance is lower. That is not a growing backlog cushion under the $8.1 billion guide. Systems YoY on this grid is +48.3%, +27.8%, +22.9%, +23.6%. CSBG YoY is +0.1%, +13.5%, +25.3%, +42.6%. China mix versus the year-ago quarter widened for three quarters and then narrowed 900 basis points in June. FY2025 columns are blank on those YoY rows because FY2024 levels are not on this grid.

The September guide is unchanged since July 29. Bettinger opened Citi on September 9 with the line that he was not going to say anything new, and he opened Goldman on September 10 the same way. Confidence is high on the direction of demand and on incremental operating margin, and only moderate on the revenue midpoint. The tone versus the prior three calls is more bullish on 2026–2027 demand, SAM, packaging, and the multi-year margin framework, and more careful about near-term upside. The clean-room constraint has not been retired. It was partially relieved, which is how $135 billion became the low-$150 billion range, and it is still why they will not say "upside bias" again.

What they will say again, and what they will not number. Calendar 2026 is the third year Lam expects to outgrow WFE, with SAM already around 36% to 36.5% of WFE, inside the high-30s destination from the early-2025 Investor Day and ahead of that timetable. At Citi, Bettinger restated the opportunity as probably north of 36%. Each 100,000 wafer starts of gate-all-around expands Lam's SAM by about $1 billion, with backside power a further $1 billion per 100,000 starts when it is adopted. The old $1.5 billion cumulative dry-resist case should be higher. He did not give the new number. Advanced packaging is on track for greater than 70% growth this year, walked up from greater than 40% in January and greater than 50% in April. Customer support stays in the June profile: NAND upgrades, spares on high utilization, and Equipment Intelligence. He would not give a September service dollar. Over the next several years they intend to drive gross margin to the mid-50s and operating margin to the mid-40s. That is several years of price, new products, and Malaysia scale, not a September number. They also said they are already ahead of the 2025 Investor Day profitability objectives. Management's own full-year scoreboard was fiscal 2026 revenue of $23.2 billion, gross margin of 50.6%, and diluted EPS of $5.82, up 41% from fiscal 2025. The four reported quarters sum to the same $23.2 billion.

Withheld. No calendar 2027 WFE number. No Lam revenue or EPS number beyond September. No segment guide. No like-for-like pricing split. Lead times are "challenging." They have met urgent requests so far, and Archer said that gets harder as the second half progresses. Clean room, not demand, is the constraint that caps how much of the $150 billion can still move up. China WFE is still flat to slightly up, unchanged, and lumpy by quarter.

Assumptions under the midpoint. WFE stays in the low-$150 billion range. The extra $10 billion versus April came from customers squeezing clean-room space, not from a new greenfield wave inside 2026. Systems and customer support both grow, and DRAM, foundry/logic, and NAND all grow versus June, which is why gross margin can hold 52% even though a pure systems surge would usually dilute it. Operating expenses grow slower than revenue. Headcount ended June at about 22,400, up about 1,800 in one quarter, mostly factory, field, and R&D. The tax rate is mid-teens, not 11%. The share count is about 1.255 billion. China does not rebound enough to change the mix and does not fall off enough to miss the midpoint. The October 2025 call had a harder China assumption that has not been refreshed in dollars: the 50% affiliate rule was then expected to cost about $600 million of calendar 2026 revenue, and China was expected to be under 30% of Lam revenue for the calendar year. July commentary did not restate the $600 million. On the two reported calendar-2026 quarters, China was 34% of March revenue and 26% of June revenue.

Read-through for the print. A typical revenue surprise is about $120 million to $145 million above the midpoint, 1.5% to 1.8% of $8.1 billion, far inside the ±$400 million range. The pattern that has improved is the margin beat. They guided September gross margin flat at the level they just beat the high end to reach. That is either conservatism, or a mix warning that systems growth is fast enough to stop the gross-margin ratchet. Both can be true. Operating margin is where they put the incremental confidence. A bullish surprise requires revenue toward the top of the band, which their history does not support, or another high-end margin beat on top of a guide that stopped raising gross margin. A disappointment is more likely to come from gross margin if the systems mix is heavier than June, from the tax rate, or from any walk-back of "everything grows," than from a small miss of a revenue midpoint they have beaten by a similar dollar amount four quarters in a row.

The December 2025 quarter is the stress test inside the sample. The October 2025 guide embedded a China restriction and took gross margin down on mix and tariffs. Revenue still beat the midpoint by +2.8%, gross margin beat by +120 basis points and cleared the high end, and EPS beat by +10.4% and cleared the high end. Conservatism survived an explicit restriction. It has not been tested against a guide the street has fully absorbed.

Consensus is underwriting the July midpoint and not another high-end beat. FMP's path after September is $8.53 billion / $2.31, $8.83 billion / $2.42, and $9.26 billion / $2.58. Versus the reported quarters those will lap, the year-ago rates are about +60% / +82%, +51% / +65%, and +38% / +42%. Growth on the street's own path peaks in December and then cools. That is a deceleration path, not an open-ended acceleration. The open questions are whether NAND's year-ago rate finally inflects, and whether 2027 WFE stays unquantified.

Management's positive list for 2027 is the fab pipeline and the technology intensity, not a revenue number. The street's September model has caught up to the only number management did give. That is agreement, not a disconnect. The disconnect is NAND's year-ago math, customer support (where the trajectory is unambiguous and the street has no line to record it), and the outer year. Deferred revenue is still below both the year-ago June quarter and the year-ago September quarter of $2,770 million, so the down-payment story is a rebuild, not a new order record.

Peer and customer news since July 29 confirms demand and does not raise Lam's official number. Micron on September 30 is the memory read. ASML on October 14 is the equipment pre-read, and that date is still an estimate. TSMC on October 15 is the foundry and packaging read. A miss, or a cut to the $8.1 billion guide, will be on the tape before KLA, Teradyne, Tokyo Electron, and Applied Materials report.

Window is July 30 through September 24, after the FY2026 Q4 print and before this one. The July 29 release is excluded. Lam's newsroom was quiet until August 13. What followed is a capacity and capital-return sequence. It does not revise the September framework. Sell-side target churn and price moves are omitted.

Three operating facts to test on October 21, none of which is a new midpoint: whether Malaysia and the lab spend show up in shipments and gross margin without a cost overrun; whether Aether and advanced packaging are still described as ahead of the old opportunity case; and whether China is slightly up in dollars while shrinking as a share. No CEO or CFO departure, no withdrawn guide, and no company release that walks the July outlook back. A September 1 report of an NT$9.8 billion Taiwan investment is footprint, in the same direction as the lab build, not a change to the revenue or margin midpoint. Insider-sale headlines were described in secondary write-ups as pre-arranged plans and were not re-pulled as a guide cut.

Consistent beater. Non-GAAP EPS beat FMP in 12 of 12 quarters and in 4 of 4 from FY2026 Q1 through FY2026 Q4. Revenue beat FMP in 11 of 12 and in 4 of 4. The only revenue miss was FY2024 Q1, −0.9%, and that quarter was still above Lam's own midpoint of $3.4 billion. The street had sat about 3% above the company's midpoint. Versus Lam's own guide, revenue and non-GAAP EPS were above the midpoint in 12 of 12. Revenue stayed inside the high end of the range in 12 of 12. Gross margin and operating margin cleared the prior-quarter midpoint in all 12 quarters, and the latest gap is the widest: gross margin +150 basis points versus a 50.5% guide, operating margin +190 basis points versus a 36.5% guide.

Magnitude is not one story. Revenue surprise versus FMP is deteriorating inside the upcycle. The last three prints stepped from +2.1% to +1.6% to +0.9%. June revenue of $6,722 million versus a $6,665 million FMP estimate is the smallest revenue beat since the FY2024 Q1 miss. FMP had already been set about 1% above the company's midpoint, so most of the guide beat was used up before the print. EPS surprise is not deteriorating. The 12-quarter average is +7.0% and the last four average +6.9%. Dollar beats rose from about $0.06 in the first four quarters to $0.10 in the last four. The last three dollar beats are +$0.10, +$0.11, and +$0.13. The percent beat on those three (+8.5%, +8.1%, +7.7%) is slightly narrower only because the denominator is rising. The air pocket was FY2026 Q1, at +3.3% and +$0.04, which is also the one recent quarter that stayed inside the EPS guide. The reason EPS keeps clearing a bar that revenue only nudges is the margin guide.

FY2024 EPS in the source system is pre-split. Lam's 10-for-1 split was effective October 2, 2024. The heatmap uses split-adjusted EPS so it matches FMP. The cited links for FY2024 Q1 through Q4 are the pre-split prints: $6.85, $7.52, $7.79, $8.14. FY2024 Q3 FMP stored the estimate as $7.30, the pre-split consensus. Split-adjusted, that estimate is $0.73 against a $0.78 actual. Using $7.30 against a split-adjusted actual would fake a miss.

Green is a beat, anchored on #1E8449. The single red cell is #C0392B. Read left to right. The revenue row darkens into FY2025 Q4 and then fades. The last revenue cell is the palest beat in the window. The EPS row stays green and re-darkens over the last three prints. No EPS cell is red. FY2027 Q1 is not on the grid.

Two EPS beats were not clean operating beats. FY2025 Q4 cleared the top of the guide with a tax rate of 4.8% on a reserve release. FY2026 Q3's 9.2% rate was tied to equity-compensation vesting. FY2026 Q4 is the clean contrast: the tax rate was 11%, in line, and the EPS beat came from gross margin. Bettinger has already told the street not to underwrite that 11% rate for September.

Read-through. A beat is still the base case on the company's own midpoint. The magnitude should not be copied from the last three EPS clears. Those three cleared the top of a range whose gross-margin midpoint sat below the quarter Lam had just printed. This guide starts gross margin at the 52% just reported. FMP is 0.4% above the revenue midpoint and about 1.3% above the EPS midpoint, and $0.12 below the EPS high end of $2.30. A print at the $8.1 billion midpoint would miss today's $8.133 billion FMP revenue estimate by about 0.4%. Lam has not printed at or below its revenue midpoint in three years. The EPS beat survives if gross margin again clears 52% by the 50 to 150 basis points it has cleared recent guides. It gets small if they merely land on the 52% and 39.5% midpoints they just handed the street.

The first session after the July 29 after-close release, July 30, closed +18.0% versus the July 29 close (Daloopa adjusted OHLCV, company_id 123, no fundamental id). That reaction is history. It is not an options-implied move for October 21. No isolated straddle for this print was sourced, so the implied-move tile is n/a.