Lam Research FY2026: The $23.2B Machine That Etches Every AI Chip

Lam Research just closed the best year in its history — revenue up 26% to $23.2 billion, net income up 36% to $7.3 billion, a June quarter that set records for revenue, operating margin and earnings per share — and then guided the September quarter to $8.10 billion, a fifth above the record it just set. No wafer fab on earth runs without Lam's deposition and etch tools, which makes this print the closest thing the AI buildout has to a capacity gauge. Every figure below is asserted line by line in the public Beancount ledger embedded at the end of this post.

The Headline Numbers​

Lam's fiscal year ends the last Sunday in June, so FY2026 is the twelve months ended June 28, 2026. Every figure below comes from the Form 10-K and the July 29 earnings release; annual history is from the FY2022–FY2025 Forms 10-K. Lam reports in thousands, and the ledger keeps all three decimals.

Read the growth rates top to bottom and the operating leverage jumps out: revenue grew 26% while cost of goods sold grew only 22% and operating expenses grew 15%, so 328 basis points of operating margin fell out almost arithmetically. This is the cleaner version of the story most equipment makers tell — Lam needed no restructuring benefit (the FY2023–FY2024 plan is finished and booked nothing this year) and no one-off gain to get there. The one line growing faster than revenue is the tax bill at +66%, as the effective rate rose from 10.1% to 12.1% — the only real drag on a print where net income still compounded faster than sales.

The June quarter itself was the exclamation point: revenue of $6.72 billion, up 30% year over year and 15% sequentially, at a 51.7% gross margin and a 37.4% operating margin — with diluted EPS of $1.81, up 25% on the March quarter. The release calls out record revenue, operating margin and EPS together, a clean sweep of the metrics management chose to highlight.

Revenue Deep Dive​

Lam disaggregates revenue two ways that matter — systems versus customer support, and the end market the equipment serves — and both cuts point at the same customer: the foundry.

Systems — new leading-edge deposition, etch and clean tools — grew nearly 30%, adding $3.39 billion that the 10-K attributes "primarily" to "foundry equipment customer spending." That is the AI-capex line in its purest form: every new fab bay needs Lam tools before it ships a single wafer. But do not sleep on the support line. Spares, services, upgrades and the Reliant non-leading-edge tools grew 20% to $8.35 billion, and the 10-K pins that "mainly" on "revenue from spares and non-leading-edge equipment." A support business growing 20% means the installed base is running hot — fabs do not buy spares for idle tools.

The end-market mix confirms where the heat is:

Foundry went from under half of equipment revenue to a clear majority in one year, on what the filing calls "mature node spending as well as investments in leading-edge equipment." Memory ceded 300 bps of share against a total that grew more than a quarter. Logic/IDM nearly halved as a share — the one soft spot, and consistent with an industry where the leading-edge logic buildout is concentrated in very few hands.

Geographically, the map barely moved at the top: China 34%, Taiwan 22% (up from 19%), Korea 19% (down from 22%), Japan 9%, the United States 7%, Southeast Asia 6%, Europe 3%. China at more than a third of revenue two years running is the number the Bear Case below is built on — every export-control headline lands directly on this line.

Management signal scan (July 29 earnings release and FY2026 10-K):

• Robust demand: "Fiscal year 2026 revenue increased 26.0% compared to fiscal year 2025, driven by strong customer demand for semiconductor equipment systems, particularly from customers within the foundry market segment" (10-K, MD&A). The release's version, from CEO Tim Archer: "Lam delivered record revenue, operating margin and earnings per share in the June quarter as AI-driven demand continues to reshape the semiconductor industry." The ledger backs both: systems revenue +29.5%, receivables up 58% on shipments surging into quarter-end.
• Outperformance framing: "Our strategic investments and technology leadership are helping customers accelerate through rising manufacturing complexity, positioning Lam for a third consecutive year of outperformance in 2026" (Archer, release). "Third consecutive year of outperformance" is a claim about wafer-fab-equipment market share, not just growth — Lam is saying it grows faster than the industry, again.
• Guidance as a demand signal: revenue of "$8.10 Billion +/- $400 Million" for the September quarter (release). A midpoint 20% above the record quarter just reported is management asserting demand visibility most cyclicals would not dare.
• Absence as a finding: there is no "demand exceeds supply" language, no upcycle claim, no pricing-momentum language, and no product-ramp boast anywhere in the release or the MD&A. The margin story is "favorable customer mix," not price increases — Lam is attributing its best margins ever to who it sold to, not to charging more. And for all the AI rhetoric, the word "backlog" never appears. The bullishness is all in the numbers and the guide, not the adjectives — which, for this management team, reads as discipline rather than modesty.

The Margin Story​

Four years of margins show a company that took a downturn gross-margin hit in FY2023, then expanded every line for three straight years:

The FY2026 gross-margin mechanics are spelled out in one MD&A sentence worth quoting whole: "Gross margin as a percentage of revenue increased in fiscal year 2026 compared to fiscal year 2025 largely due to favorable customer mix, partially offset by aluminum and steel tariff-related spend." Two things matter here. First, mix — foundry tools evidently carry better margins than the average sale, so the 900-basis-point foundry shift did the heavy lifting. Second, tariffs are already a measured headwind inside cost of goods sold, not a hypothetical risk-factor item; gross margin expanded despite them, which bounds how bad the tariff drag currently is but also means any escalation comes straight out of this line.

Below gross profit, the opex story is headcount. "The increase in operating expenses in fiscal year 2026 compared to fiscal year 2025 was primarily due to employee-related costs from increased headcount and higher supplies spending for research and development." R&D headcount growing into a 35%-margin year is exactly what a technology-moat bull wants to see — Lam spent $2.38 billion on R&D, 10.2% of revenue, while expanding operating margin anyway. SG&A at 4.9% of revenue remains almost comically lean for a $23 billion company; selling wafer-fab tools to a customer base so concentrated that four buyers owe over half of receivables takes no sales army.

The net-margin bridge has one headwind worth naming: the effective tax rate rose two full points to 12.1%, adding nearly $400 million to the tax line. At the FY2025 rate, net income would have printed roughly $160 million higher. Even taxed harder, Lam kept 31 cents of every revenue dollar — a higher net margin than most software companies manage.

The One Big Question: Can Lam Actually Ship the $8.1 Billion September Quarter?​

The September guide deserves its own section because of its sheer size. A $8.10 billion midpoint is 20% sequential growth on top of a record — more in ninety days than half of what the entire FY2024 trough year produced. Three checks decide whether to believe it.

Check one: the backlog proxy. Lam does not disclose backlog, but deferred profit — revenue less related costs on shipped tools awaiting acceptance — is the closest ledger line, and it fell $286 million year over year to $2.28 billion even as revenue grew 26%. A purist reads that as demand digestion, not accumulation. The counter is timing: the release says deferred revenue rose from $2.22 billion to $2.43 billion within the June quarter alone, and another $490 million of Japan shipments sits in finished-goods inventory awaiting customer acceptance. The annual drawdown with a quarterly rebuild says Lam shipped everything it could all year and started refilling the pipe in June. Verdict: consistent with the guide, but this is the line to watch in September — if deferred balances fall again while revenue hits $8.1 billion, Lam is eating its seed corn.

Check two: the receivables bill. Accounts receivable nearly doubled the pace of revenue — up 58% to $5.34 billion while sales grew 26% — pushing days-sales-outstanding from 67 to 84 days. The cash-flow statement shows the damage: a $1.96 billion receivables build that left operating cash flow down 5% in a year net income rose 36%. Some of this is just June-quarter linearity — a record quarter that back-loads shipments always strands receivables at year-end — and the filing says the doubtful-accounts allowance is immaterial, so Lam does not expect these bills to go unpaid. But four customers owe 52% of that $5.34 billion (16%, 15%, 11% and 10% each, per the concentration note), which means a handful of payment calendars are Lam's cash flow. If September's $8.1 billion ships on similar terms, operating cash flow will undershoot earnings again, and the "show me the cash" crowd gets louder.

Check three: the only peer that matters. Applied Materials — the other American equipment giant, reporting a July quarter that overlaps Lam's June quarter almost exactly — printed $9.12 billion of revenue at a 50.3% gross margin and 33.7% operating margin in its July 2026 quarter. Line them up:

Lam is smaller and more profitable on every line — 140 bps of gross margin and 370 bps of operating margin ahead of its larger rival in effectively the same demand environment. That gap is the mix story again: Lam's etch and deposition tools are riding the foundry wave harder than AMAT's broader portfolio this cycle. It also frames the $8.1 billion guide differently — Lam is not guiding into a vacuum; it is guiding to close the absolute-revenue gap with AMAT while holding better margins. If both companies hit their guides, the equipment duopoly debate gets very interesting.

Tracking a $23 Billion Company in Plain Text​

A $23 billion equipment maker fits in a plain-text ledger because double-entry bookkeeping forces every dollar to reconcile — revenue minus every expense must equal net income to the thousand dollars, or the file refuses to balance. That discipline is the point: this post's tables are not screenshots of a spreadsheet but assertions a computer re-verifies. Lam's chart of accounts is modeled on how we model every company, with the equipment-maker shape it shares with Applied Materials — an explicit Inventories line, and Lam's "Deferred profit" caption mapped to DeferredRevenue.

Here is FY2026's income statement exactly as the ledger records it (income is negative because it is a credit; the check comment above the transaction proves the postings sum to zero):

; FY2026 Income Statement — fiscal year ended June 28, 2026
; Check: −23232.690 + 11507.382 + 2375.873 + 1149.640 − 62.678 + 997.077 + 7265.396 = 0 ✓
 
2026-06-28 * "Lam Research Corporation" "FY2026 Income Statement"
  Income:Revenue                           -23232.690 MUSD
  Expenses:CostOfRevenue                    11507.382 MUSD
  Expenses:ResearchAndDevelopment            2375.873 MUSD
  Expenses:SellingGeneralAdministrative      1149.640 MUSD
  Income:OtherNet                             -62.678 MUSD  ; other income, net
  Expenses:IncomeTax                          997.077 MUSD
  Equity:Adjustments                         7265.396 MUSD  ; reported net income offset

The balance-sheet number that tells the FY2026 narrative is receivables: $3.38 billion to $5.34 billion in one year, the $1.96 billion build that swallowed the entire net-income gain before operating cash flow. Flip to the other side of the ledger and the capital-return machine is just as loud — $3.85 billion of buybacks plus $1.27 billion of dividends consumed 87% of operating cash flow, and treasury stock now exceeds paid-in capital more than threefold, leaving the common-equity account in a net debit. Lam is, balance-sheet-wise, a company that has bought back more than it ever raised and still sits on $1.84 billion of net cash. Browse all five years below:

The Multi-Year Arc​

Five years of the ledger compress the semiconductor cycle into one table: the FY2024 trough, when revenue fell 14% and the company booked its last restructuring charges, followed by two years of 23.7% and 26.0% growth that lifted the top line 56% from the bottom.

Three things stand out. First, the trough never broke the machine: even in FY2024's down 14%, Lam held a 28.6% operating margin and a 25.7% net margin — the support business ($5.98 billion of spares and services at the very bottom, down only 11% while systems fell 17%) is a shock absorber most cyclicals lack. Second, per-share compounding is doing quiet work the revenue line hides: the October 2024 ten-for-one split restates history, but the share count keeps falling underneath it (1.29 billion diluted shares to 1.26 billion this year alone), which is why EPS grew 39% against net income's 36%. Third, the cash line is now the odd one out — operating cash flow compounded slower than earnings for two straight years as working capital absorbed the growth. In FY2025 a $1.15 billion deferred-profit inflow papered over the receivables build; in FY2026 there was no such offset, and the $1.96 billion receivables build stands exposed. The arc says this company converts eventually. FY2027 has to prove it.

The Verdict: Bull vs. Bear​

Bull Case

• The guide is the thesis. $8.10 billion for September (+20% sequential) with a 39.5% operating margin says AI-driven foundry demand is accelerating, not peaking — on top of a June quarter that itself grew 15% sequentially and 30% year over year.
• Foundry mix is margin mix. The 900-basis-point shift to 54% foundry drove 176 bps of gross-margin expansion via customer mix; every incremental leading-edge fab bay favors Lam's etch and deposition tools.
• Support revenue is the floor. $8.35 billion of spares, services and Reliant tools (+20%) means the installed base is sweating — recurring-ish revenue that held at $5.98 billion even in the FY2024 trough.
• Capital return without leverage. $5.12 billion returned (87% of operating cash flow) while holding $1.84 billion of net cash and repaying the $750 million 2026 notes — the balance sheet funds both the buyback and the growth.
• Best margins in the duopoly. 37.4% quarterly operating margin beats Applied Materials' 33.7% in the same demand window — Lam extracts more profit per revenue dollar than its larger rival right now.

Bear Case

• China is a third of revenue. 34% of sales into the jurisdiction every export-control headline targets; the release's own risk list names trade regulations, export controls and geopolitical tensions as threats to exactly this outlook. One rule change rewrites the guide.
• Receivables say the growth is on credit. DSO up 17 days to 84, a $1.96 billion build that pushed operating cash flow down 5% while earnings rose 36% — if September ships the same way, cash conversion stays broken.
• $8.1 billion leaves no room. A 20% sequential guide into tariff headwinds "partially offsetting" gross margin and a two-point higher tax rate means any slip — a delayed fab, a pushed tool acceptance — misses loudly.
• Deferred profit is not confirming. Down $286 million year over year at the same time revenue grew 26% — the backlog proxy shrank in a record year. The quarterly rebuild helps, but one quarter is not a trend.
• The cycle always collects. FY2024 was two years ago: revenue down 14%, restructuring charges, trough margins. Equipment orders lead wafer demand by quarters, and today's 54% foundry mix concentrates Lam in exactly the customers whose capex cuts deepest when memory turns.

Our Take: Believe the guide but watch the cash. Lam earned its record year the hard way — mix-driven margin expansion, no one-offs, R&D up 13% into strength — and the foundry concentration investors fear is precisely what is paying them 37% operating margins today. But a company that grows net income 36% while operating cash flow falls is asking to be paid on delivery, not on shipment. If September's $8.1 billion converts receivables back into cash, this is the best equipment franchise in the cycle. If DSO pushes past 90 days, the growth is being bought with terms. The ledger will show it first — that is what it is for.