LRCX — Lam Research Corporation

Lam Research controls roughly 40% of the global plasma etch equipment market through switching costs so entrenched that customers require 12 to 18 months and tens of millions of dollars to qualify a competitor's tool — a structural barrier that produces $6.9 billion in annual recurring service revenue growing even as semiconductor capital expenditure contracts. The business is genuinely excellent: gross margins have expanded from 46% to 50% over five years, free cash flow runs at $5 to $6 billion annually, and the AI-driven buildout of advanced logic and memory capacity is structurally real. But at 55 times trailing earnings — nearly three times the company's own ten-year historical average multiple — the stock prices in every one of these advantages in full, leaving no room for the cyclicality that sent revenue down 14.5% just two fiscal years ago.

---

The semiconductor equipment industry spent 2022 and 2023 reminding investors that it is, structurally excellent or not, a cyclical business. Memory producers — the largest customers for equipment spending — responded to post-pandemic demand destruction by cutting capital expenditure sharply. Wafer fabrication equipment orders collapsed. Lam Research's revenue fell from $17.4 billion in fiscal 2023 to $14.9 billion in fiscal 2024, a 14.5% decline that management characterized, in the same breath, as a temporary dislocation within a durable secular growth story. They were largely correct: by fiscal 2025, revenue had recovered to $18.4 billion; and for the quarter ending March 2026, the company generated $5.84 billion in a single quarter — annualizing to a revenue run rate approaching $23 billion — with operating margins of 35%. The equipment oligopoly had not only recovered; it had re-accelerated, carrying with it a valuation that now prices the recovery and several years of future growth simultaneously.

In early 2026, the consensus narrative around semiconductor equipment is compelling enough to have driven a significant re-rating across the entire sector. Wafer fabrication equipment spending is projected at $133 billion in calendar 2025, rising toward $145 billion in 2026 and a projected record of $156 billion in 2027. TSMC is guiding capital expenditure of $52 to $56 billion in 2026, up 27 to 37% from the prior year. Samsung, SK Hynix, and Micron are each expanding DRAM capacity to feed demand for high-bandwidth memory — the specialized chips that make AI inference economically viable at scale. Every advanced chip produced in this buildout passes through equipment made by one of five dominant companies: ASML, Applied Materials, Lam Research, Tokyo Electron, and KLA Corporation. These five collectively control more than 60% of global equipment spending, and the barriers protecting their positions have been built over decades of accumulated process expertise, customer relationships, and proprietary technology.

Semiconductor equipment is structurally unlike most capital goods industries. The know-how required to precisely etch patterns at sub-5 nanometer geometries — removing atoms one layer at a time without disturbing structures immediately adjacent — is not approximatable by a well-funded new entrant. ASML spent more than $10 billion and over twenty years developing extreme ultraviolet lithography before shipping a production-capable system; no competitor has come within a decade of threatening that position. The switching dynamics are equally protective for equipment already deployed: when a chipmaker qualifies a piece of equipment on their production line, that qualification involves months of process characterization, recipe tuning, and yield verification. Replacing a qualified tool requires restarting that process from scratch. This is not a product category where the lowest bid wins the next purchase order.

Lam Research operates in two segments with fundamentally different economic characteristics. The first is systems — new leading-edge equipment for deposition, etch, clean, and advanced packaging processes. In fiscal 2025, systems revenue was approximately $11.5 billion. The second is the Customer Support Business Group, or CSBG — covering spare parts, equipment upgrades, service contracts, refurbished tools, and lifecycle support for the 100,000-plus chambers already installed in customers' fabs. CSBG generated $6.9 billion in fiscal 2025, and — this is the analytically important number — held at $6.6 billion in fiscal 2024, even as total revenue fell 14.5%. Systems revenue swings with capex cycles; CSBG compounds with the installed base. Every chamber Lam places in a customer fab generates CSBG income for 15 to 20 years. The installed base has become an annuity. Management has set a target of growing CSBG to 1.5 times its 2024 level by calendar 2028 — from roughly $6.6 billion toward approximately $10 billion — without requiring a single incremental new systems sale to achieve it.

The foundation of Lam's competitive position is its dominance in plasma etch equipment. Etching — selectively removing material from a wafer to define the microscopic structures that make up a chip's circuitry — is one of the most technically demanding steps in semiconductor fabrication. Lam holds approximately 40% of the global etch market. In the specific application of high-aspect-ratio etch for 3D NAND flash memory, the company's share approaches 70 to 80%. High-aspect-ratio etch for NAND requires cutting precise columns through stacked layers 60 or more times taller than they are wide — the precision equivalent of drilling a perfectly straight needle through a foot of granite — and Lam has maintained better results than competitors across a decade of NAND scaling from fewer than 100 layers to 200 layers and beyond. The Akara platform, launched in February 2025, targets conductor etch for gate-all-around transistors, extending the etch leadership into the next generation of logic device architectures.

The commercial depth of the moat exceeds even the technical advantage. When a chipmaker qualifies a Lam etch tool on their production line, the recipe — the precise sequence of plasma chemistry, timing, temperature, and pressure — is embedded into the manufacturing process at the atomic level. Switching to a competitor's tool does not simply require buying new equipment. It requires qualifying a new recipe from scratch, a process that takes 12 to 18 months and costs tens of millions of dollars per customer. No memory producer facing a competitive environment, a device ramp, or a product transition does that voluntarily. The margin data confirms the pricing power this creates:

Company	Gross Margin	Operating Margin	Service Revenue %	Etch Position
Lam Research	50.6%	35%	43%	~40% global; ~70–80% NAND
Applied Materials	48.7%	~25%	23%	Smaller; gaining in metal etch
Tokyo Electron	47.6%	~28%	~25%	15–20%; growing in logic etch

Lam's 50.6% gross margin exceeds both Applied Materials and Tokyo Electron despite a more concentrated product portfolio. Its 43% service revenue share — nearly double Applied Materials' 23% — reflects the compounding value of the installed base. The operating margin gap is more striking still: Lam at 35% versus Applied Materials at approximately 25%. This is not a scale advantage; Applied Materials generates more total revenue. It is pricing power and revenue mix, both of which trace directly to the depth of customer lock-in. The moat is measurable, not asserted.

It is not, however, unassailable. Applied Materials has been investing specifically in metal etch technology for advanced interconnects — the application where Lam's lead is narrowest — and has been closing the gap at the margin. Tokyo Electron is growing its logic etch position in Asia-Pacific markets. Chinese domestic equipment makers, primarily Naura and AMEC, have captured roughly 35% of domestic Chinese market share in mature-node etch, though they operate exclusively at 28 nanometers and above, with no presence in the sub-5nm applications that define Lam's highest-value work and no access to the restricted technology that underlies it. The moat is wide in its core application; it is narrower in certain adjacent applications where the competitive picture is more contested.

In fiscal 2025, Lam generated $18.4 billion in revenue, up 23.7% from the FY2024 trough. Gross margin was 48.7% — the highest level since 2012. Free cash flow was $5.4 billion, or 29% of revenue. For the quarter ending March 2026, gross margin reached 49.9% and operating margin was 35%; the company guided the following quarter at $6.6 billion in revenue, 50.5% gross margin, and 36.5% operating margin. GAAP and non-GAAP earnings per share differ by less than $0.02 per quarter — the reconciling items are amortization of a small amount of acquired intangibles and debt discount, totaling roughly $3.5 million per quarter. The reported numbers can be taken essentially at face value. One item worth naming plainly: China revenue represented approximately 42% of total in fiscal 2024, has declined to approximately 34% as of the March 2026 quarter, and management expects continued normalization below 30% as new export control restrictions take effect. The estimated revenue impact of those restrictions is $600 to $900 million cumulatively — a real headwind that the recovery narrative somewhat papers over.

Tim Archer has led Lam since January 2018. Under his tenure, gross margins have expanded from roughly 44% to 50%, free cash flow has grown from approximately $2 billion to $5.4 billion annually, and CSBG has been developed into a structural differentiator rather than a supplemental revenue line. The capital allocation record is aggressive: over the trailing twelve months through the March 2026 quarter, Lam repurchased $3.3 billion in shares against a $10 billion authorization announced in May 2024. In the most recent quarter, the company returned $800 million in buybacks, $326 million in dividends, and retired $750 million in debt — 139% of free cash flow. The company has maintained an uninterrupted dividend for twelve consecutive years. Archer's direct stake — approximately 0.065% of outstanding shares, worth roughly $218 million at current prices — is meaningful in absolute terms but modest for an eight-year CEO tenure at a company of this scale. Compensation is 95.6% variable, linking rewards to stock performance. The capital return record speaks most clearly: this management team deploys cash at high rates and does not pursue empire-building through dilutive acquisitions.

The growth runway table tells two stories at once. The first is cyclicality — revenue peaked, fell 14.5%, and recovered. The second is structural improvement — CSBG held its ground through the trough, gross margins expanded through both boom and bust, and free cash flow remained substantial even in the down year.

Fiscal Year	Revenue	CSBG Revenue	Gross Margin	Free Cash Flow
FY2022	$17.2B	—	45.7%	~$3.9B est.
FY2023	$17.4B	—	45.1%	$4.7B
FY2024	$14.9B	$6.6B	47.6%	$4.3B
FY2025	$18.4B	$6.9B	48.7%	$5.4B
FY2026 (run rate)	~$25B	~$8.4B	~50%+	est. ~$7B+

Note: CSBG was not separately disclosed in widely available filings prior to FY2024; FY2026 figures annualize Q3 FY2026 actuals ($5.84B quarter) and Q4 FY2026 guidance ($6.6B quarter); Lam's fiscal year ends in late June.

The margin expansion across this table is structural, not cyclical. In FY2023, at $17.4 billion in revenue, gross margin was 45.1%. In FY2024, at the $14.9 billion trough, gross margin was 47.6% — higher than the prior peak-revenue year. In FY2025, revenue recovered and gross margin reached 48.7%. The current run rate is approaching 50.5%. The gains reflect CSBG's growing share of total revenue, pricing discipline enforced by switching costs, and factory efficiency improvements. They are not going to reverse when the next cycle softens.

The penetration argument for the growth runway rests on two mechanics. The first is the NAND layer count trajectory. 3D NAND flash memory has been scaling aggressively, with stacked layer counts growing from under 100 in 2018 to 200 today, targeting 300 layers within the next several years. Each additional layer requires additional etch steps — Lam's content per wafer grows linearly with the layer count even at flat NAND unit shipments. Lam's share of the NAND equipment market is expanding from approximately 30% in fiscal 2024 toward 40 to 45% in fiscal 2025, with management projecting capture of more than two-thirds of the estimated $40 billion in NAND upgrade equipment spending that remains ahead through calendar 2027. The second mechanic is the installed base compounding into CSBG. With more than 100,000 chambers in service — each generating 15 to 20 years of recurring service revenue — the current installed base alone supports the targeted 1.5× CSBG growth by calendar 2028. At an average annualized CSBG contribution of roughly $70,000 per chamber across the installed base, the existing 100,000 chambers imply approximately $7 billion in structural annual CSBG capacity without adding a single new tool. Systems growth is upside; CSBG compounding is the floor. The company has captured the dominant position in the single most etch-intensive end market in semiconductor manufacturing. What remains uncaptured is the next decade of NAND layer scaling and the advanced packaging expansion — a market where Lam's revenue more than doubled in calendar 2024 and is projected to grow more than 50% in calendar 2026.

As of mid-April 2026, Lam Research's stock trades at approximately $272 per share, implying a market capitalization of approximately $334 billion and an enterprise value of approximately $332 billion. The trailing GAAP price-to-earnings ratio is 55 times. Using management's current guidance trajectory, the forward multiple is approximately 43 times. The company's own ten-year historical average price-to-earnings ratio is 19 times. The stock is trading at nearly three times its historical average. Compared to peers: Applied Materials trades at approximately 41 times trailing earnings and 33 times forward earnings; ASML at 49 times trailing and 37 times forward; KLA Corporation at approximately 52 times trailing. Lam is the most expensively valued of the major equipment companies on a trailing basis and clustered at the top on a forward basis. EV/EBITDA of 45 times compares to 35 times for Applied Materials and 38 times for ASML.

The bull argument for the current valuation has three elements that deserve honest treatment. First, the AI capex cycle is not a two-quarter phenomenon — the buildout of inference infrastructure is multi-year, driven by hyperscalers who have made public capital commitments that are difficult to retract quickly. Second, CSBG's growing share of total revenue structurally raises the earnings floor, making the historical average multiple — computed in an era when CSBG was a smaller portion of the business — partially obsolete as a reference point. Third, the NAND layer count trajectory means Lam's content per wafer grows with the advance of the technology, independent of unit volume. These are real arguments, not rationalization.

The bear case is equally specific. A 14.5% revenue decline occurred in fiscal 2024 — not in a prior decade, but two years ago — in a period when the AI narrative was already well established and chipmakers were already invested in the multi-year capacity buildout story. Equipment cycles do not respect secular narratives when chipmakers are sitting on inventory or margins are compressing. At 55 times trailing earnings, even a modest pause in NAND capex — the kind that happens roughly every three years — would compress earnings meaningfully at a moment when the multiple offers no cushion. A re-rating from 55 times to 35 times, combined with 20% earnings growth over two years, still produces a flat-to-negative stock return from today's price. The investor at $272 is not paying for a business with a margin of safety; they are paying for a business where every favorable assumption must come true to avoid loss.

The most precise statement of the bear case is this: the CSBG floor — the strongest structural argument for a re-rating relative to historical averages — is already visible in the trailing earnings. The $6.6 billion CSBG contribution held fiscal 2024 gross margin at 47.6% rather than the 40%-range results of prior troughs. The trailing earnings of $4.89 per share for fiscal 2025 already incorporate the structural improvement CSBG provides. Applying 55 times to that number is not paying a premium for future improvement not yet in the earnings — it is paying a premium for present improvement that is fully recognized, plus an additional premium for future improvement that is not yet demonstrated. The answer is partial: CSBG will be higher in the next cycle than in fiscal 2024, because the installed base grows with every new systems placement. But the increment is incremental, not transformational, and it is being paid for at a rate that leaves no room for any of the things that have historically interrupted equipment cycles.

For the investment case to change to a positive verdict, one of two things would need to happen. Either the stock would need to decline meaningfully — at approximately 30 times forward normalized earnings, in the range of $130 to $150, the price begins to embed realistic rather than optimistic assumptions and the earnings yield generates a positive real after-tax return — or the company would need to deliver several consecutive years of earnings expansion at a rate that validated the current multiple as reflecting fair value rather than cycle enthusiasm. Until one of those conditions arrives, the outstanding quality of the business is fully priced and no additional return accrues to the investor who pays today's price.

The etch process is irreplaceable. The installed base is real and growing. The free cash flow is genuine. The stock, at nearly three times its own historical average, is not reasonable.