GARPify
We do the work. You do the thinking.
LRCX|Story|Leadership|What’s ChangedEarnings|Fin. Strength|Valuation|Risk|Sentiment|Technical|Think|Disclaimers
Lam Research Corporation
NASDAQ · LRCX  ·  $314.00  ·  August 21, 2026  ·  Q4 FY2026
This Report
August 21, 2026
Q4 FY2026 · GARPify
Next Earnings Release
Late October 2026
Q1 FY2027 · Lam Research Corporation IR
Next GARPify Update
Q1 FY2027
Following earnings · subject to change
GARPify theme · where this report begins
The Machine Behind Every Chip
Chipmakers compete; their equipment suppliers collect either way. Etch and deposition — sculpting the structures themselves — sit at the heart of that spending, and Lam is one of its two indispensable names.
This company’s role
Lam is the theme’s sculptor — its etch and deposition machines carve and build the three-dimensional structures that modern chips are made of, layer by layer, at every leading fab. The theme explains why the wave exists; this report asks whether Lam at today’s price is the way to ride it. Read the theme →
Over time, a stock is a weighing machine — and it weighs earnings. This report weighs LRCX’s, then asks the four questions that turn a great company into a good investment.
Question 1
Is the wave real?
FY26 revenue $23.2B · +26%every 3-D structure is etched & depositedSept quarter guided to $8.1B
Yes. Modern chips are no longer printed flat — they are built, layer by carved layer, in three dimensions, and the carving and layering is what Lam’s etch and deposition machines do. The artificial-intelligence build-out multiplies exactly that work: high-bandwidth memory stacks dozens of storeys high, advanced logic wraps transistors in new geometries, and every added layer passes through Lam’s chambers. Fiscal 2026 revenue reached $23.2B, up 26%, and management just guided the September quarter to $8.1B ± $400M — a step-up of a fifth in one quarter (accession 0000707549-26-000033). The wave is real, and Lam is one of the two names it cannot route around.
Question 2
Is this a great company?
EPS 27%/yr over 10 yrsgross margin 50% · operating 35%49% return on money invested · shares -22% in 10 yrs
Yes — a great cyclical-growth company, with the emphasis on both words. Earnings per share compounded 27% a year for a decade ($0.52 to $5.76, split-adjusted) on 50% gross margins, 35% operating margins and a 49% return on invested capital, while buybacks retired 22% of the shares. The cycle is the asterisk: revenue fell in fiscal 2019 and again in fiscal 2024 (-14%), and earnings swing harder than revenue in both directions. The five-year compounding rate, 16%, is more honest than the ten-year figure because it starts from a boom year rather than a trough. Quality: emphatic. Smoothness: not part of the deal.
Question 3
Am I paying a fair price?
54× now · 2.1× its norm5-yr: 9× low · 26× avg · 78× highavg multiple → ~$151
No — and by a wider margin than any stock in this coverage. Over five years the market has paid between 9× and 78× for Lam’s reported earnings, averaging 26×; today’s price asks 54×, roughly 2.1 times the normal toll. At the average multiple, current earnings support a price near $151 against $314 on the screen. The defense — record results, a $8.1B September guide, earnings up 39% in a year — is real, and it is also the point: paying 2.1 times the historical multiple means the record year and the next one are both already in the price. The earnings yield is 1.8% against a 4.707% Treasury; the PEG on five-year growth is 3.3 against a GARP threshold of 1. Magnificent quarter, expensive decade of assumptions.
Question 4
Is the trend my friend?
$314 · +11% vs 1-yr avg52-wk $97–$433-28% off the June peak
Yes. On the rule this framework uses, price sits about 11% above a 1-year average that is still rising. And the context matters just as much here as at its theme-mate: the stock ran from $97 to $433 inside a year — better than a four-fold rise — then gave back 28% in the July air pocket. The average is rising steeply because that vertical leg is still inside the 30-week window. The same honest label applies: an uptrend that has just been stress-tested once, with the price about 11% above a fast-rising line.
Factor reads — level & direction
Earnings growth
39%
▲ rising
45th pctile of own history
Revenue growth
26%
▲ rising
73th pctile of own history
Return on capital
49%
▲ rising
100th pctile of own history
Operating margin
35%
▲ rising
92th pctile of own history
Free cash flow margin
27%
↔ steady
67th pctile of own history
Gross margin
50%
↔ steady
92th pctile of own history
Earnings yield
1.84%
▼ falling · below risk-free
1.84% vs 10-yr UST 4.707% · ERP -2.9 pts
Shares outstanding
1.25B
▼ shrinking
10-yr net -22% · buybacks vs. dilution
Financial strength (Altman Z)
34.1
↔ deep in the safe zone
Z-score · >3 safe, <1.8 distress
GARPify — An interactive approach to investment selection.
Rate each section as you read
and see your results on the Think tab.
Start with the Story  →
GARPify draws on the frameworks of seven investors who built their track records in public
We do not invent our methodology. We stand on the shoulders of people who did the work first.
Benjamin Graham
The weighing machine · price vs. worth · margin of safety
The Intelligent Investor →
Peter Lynch
GARP · find the compounder before the crowd
One Up on Wall Street →
Warren Buffett
Business quality · moat · long-term ownership
The Essays of Warren Buffett →
Charlie Munger
Mental models · invert · think clearly
Poor Charlie's Almanack →
Jim Collins
What makes great companies endure · flywheel
Good to Great →
Howard Marks
Cycles · sentiment · the pessimistic price signal
The Most Important Thing →
Next
The Lam Research Corporation Story →
The most important machines in the world’s most important factories do their work one atomic layer at a time, in vacuum chambers, in the dark. The company that builds them just finished the best year in its history — and its stock quadrupled, then dropped a quarter in eight weeks. The machines never changed. The price did. Deciding what that means is this report’s whole job.
The Lam Research Corporation Story
Applying the frameworks of Benjamin Graham, Warren Buffett, Charlie Munger, Peter Lynch, Jim Collins, and Howard Marks
The organizing idea. Benjamin Graham said that in the short run the market is a voting machine, but in the long run a weighing machine — and what it weighs is earnings. Everything that follows is built to weigh LRCX’s: the engine that produces the earnings, the people driving it, and the price you are asked to pay for the result.

A modern memory chip is a skyscraper you could lose on a fingertip — hundreds of storeys of circuitry, stacked and wired in three dimensions. Nobody prints such a thing. It is sculpted: material laid down one atomic layer at a time, then carved away with plasma in patterns finer than a wavelength of light. The machines that do the laying and the carving — deposition and etch — come overwhelmingly from a short list of companies, and at the top of that list sits Lam Research. Its chambers run at every leading chipmaker on earth. When any of them adds a layer, Lam gets paid.

Part 1 - History
The sculptor in the dark

The economics compound with the physics. Each generation of chips needs more layers than the last — three-dimensional memory went from dozens of storeys to hundreds; advanced logic keeps wrapping its transistors in new geometries — so etch and deposition intensity rises with every node, cycle after cycle. Over the decade that arithmetic took Lam’s revenue from $5.9B to $23.2B and earnings per share up 27% a year, while margins climbed to 50% gross and 35% operating — extraordinary levels for machines that weigh tons and cost tens of millions of dollars apiece.

There is a second business hiding inside the first. Every chamber Lam has ever installed needs spares, service and upgrades for decades, and that installed-base franchise — $8.3B of fiscal 2026 revenue, about 36% of the total (10-K accession 0000707549-26-000037) — keeps earning through the downcycles when new-tool orders pause. It is the shock absorber that let free cash flow stay near 29% of revenue even in fiscal 2024, when revenue itself fell 14%.

Part 2 - Present
The best year, itemized

Fiscal 2026 was the best year in company history, and it is worth itemizing: revenue $23.2B, up 26%; earnings per share $5.76, up 39%; gross margin above fifty percent for the first sustained stretch; $5.1B returned to shareholders; and a September-quarter guide of $8.1B that would be, by a wide margin, the largest quarter Lam has ever reported (accessions 0000707549-26-000033, 0000707549-26-000037). The artificial intelligence (AI) build-out — high-bandwidth memory above all — is pulling harder on Lam’s specific machines than any cycle before it.

The stock noticed. Twelve months ago the shares traded near $97; at the June 2026 peak they touched $433; today they sit at $314, -28% below that peak. Decompose the year and the story writes itself: price +223%, earnings +39% — the other 133 points were the multiple, expanding from the low twenties to today’s 54×. The market did not merely reward the best year in company history; it paid more than twice its own five-year-average price for each dollar of it.

Part 3 - Future
Priced for a perfect decade

That is the argument of this report in one line: at 54× reported earnings — against a 26× five-year average and a 9× downcycle low — the wonderful is fully paid for. The stock’s own history supplies the caution: this franchise saw its revenue fall twice in the past eight fiscal years, and its multiple traded below 26× for most of that period. The honest bull case is that the layer-count physics and the memory build-out have permanently raised Lam’s growth rate. The honest observation is that permanence is exactly what a 2.1-times-normal multiple requires.

What would change the arithmetic? The same two paths as ever: a price nearer the historical toll — the $151 area on current earnings, or higher as earnings grow — or a year or two of 16%-plus earnings growth at a flat price, letting the business catch up to its ticket. Neither requires anything to go wrong in the chambers. The tabs ahead grade each signal; the watch-list is a position too.

What You'll Discover
The story raises the questions. The tabs that follow answer them with data - each ends with a box for you to rate, and your conviction takes shape on the Think tab.
1
Leadership
Engineers who under-promise: four of the last five quarters delivered above the entire guided range, and a dividend raised every year for a decade.
2
Earnings Growth
Cyclical compounding at its most violent and most rewarding — 27% a year for a decade, through two downturns, to a record fiscal 2026.
3
Financial Strength
An Altman Z of 34 and 36% recurring installed-base revenue — the fortress that funds $2.4B of annual research through every winter.
4
Valuation
The reddest light in this coverage: 54× against a 26× five-year average — a record year, fully pre-paid.
5
Risk
Price, China (33.8% of revenue) and the memory cycle top the list; the franchise itself — plasma, physics, installed base — anchors the bottom.
6
Sentiment
From 9× despair to 78× euphoria in four years — the market’s wildest voting record, currently chastened but still premium.
7
Technical
A parabolic uptrend, stress-tested once: +11% above a rising 1-year average whose slope is mostly the June spike.

The bottom line.On the four-signal GARP framework, Lam reads as the same contradiction as its theme-mate, drawn larger: a wave that cannot crest (Light 1 — every new chip generation adds layers, and every layer is etched and deposited); a business of the first rank (Light 2 — 50% gross margins, 49% returns on capital, $8.3B of installed-base revenue cushioning the cycle); a price at 2.1 times its own five-year-average multiple with a 2.9-point negative spread to the risk-free rate (Light 3 — the most expensive-versus-history reading in this coverage); and a chart +11% above a rising 1-year average, eight weeks after a 28% air pocket (Light 4 — constructive, on probation). The framework’s conclusion is patience with a pencil: an extraordinary business whose price already assumes the extraordinary continues uninterrupted.

From Here
History explains how a maker of vacuum chambers became indispensable to the most strategic industry on earth. The present is the best year that company has ever printed. The future question is the uncomfortable one the last twelve months forced: whether a business this cyclical, this concentrated by geography, and this good can be worth more than twice its own average multiple. The tabs ahead put that question to the evidence, one signal at a time. Read them in order, rate each as you go, and let your own conclusion assemble itself.
Does this story resonate with you?
Your conviction:
Next
Leadership & Plan →
Who is driving the bus, how they’ve performed, and their plan.
Financial Strength
STRONG
You’ve met the driver and the engine. Can the bus take a rough road — and still climb?

A balance sheet does two jobs, and a great business needs both. It has to survive trouble — a recession, a bad year, a job that goes wrong — without the wheels coming off. And it has to be strong enough to fund the plan under its own power — paying for the growth management just laid out from its own cash, not by piling on debt or selling new shares. Defense and offense.

So this tab checks the frame. A weak balance sheet can strand even a wonderful engine the moment the road turns bad; a fortress one lets the earnings keep compounding through every kind of weather — which is the whole point, because those are the earnings the market eventually weighs. The question is simple: is LRCX built to go the distance?

"Only when the tide goes out do you discover who has been swimming naked." For a company whose customers can halve their orders inside a year — and have, twice in eight years — the balance sheet is the swimsuit. This tab asks whether Lam can sail through the next downcycle with its research budget, its dividend and its franchise intact.

Is the balance sheet a strength or a risk?
A fortress by any measure: Altman Z of 34, current ratio 2.6, $4.0B of long-term debt against $4.9B of annual free cash flow — under a year of cash generation retires all of it.
Can the company survive a downturn?
Cash conversion holds through the storms: free-cash-flow margin stayed near 29% in the fiscal-2024 downcycle because $8.3B of installed-base revenue bills regardless of the order book.
Is the business generating real cash?
Returns keep flowing: $4.9B to shareholders in fiscal 2026, a dividend raised every year of the decade (now $0.26/quarter, accession 0000707549-25-000082), and a $10.0B buyback authorization.

Read each like a price chart. The actual metric (solid line, green when rising / red when falling) is plotted with its trend line (grey dashed) on the same axis. On mean-reverting metrics like margins and returns, a tinted band also marks the level versus the metric's own ten-year median; on cumulative-dollar metrics like free cash flow that band carries no signal, so it's omitted.

Free cash flow$4.9B
▲ uptrend · actual vs trend moving average · The cash left after the business runs and grows itself.

Free cash flow was $4.9B in fiscal 2026 at a 27% margin. The decade’s pattern matters more than any single year: conversion has held between roughly 14% and 29% of revenue through booms and downcycles alike.

Free cash flow margin27%
▲ steady-high · actual vs trend moving average · How much of every sales dollar survives as cash.

The downcycle test is the one that counts, and fiscal 2024 ran it live: revenue fell 14%, and free-cash-flow margin came in at 29% — near the decade’s high — because the $8.3B installed-base franchise keeps billing when new-tool orders pause. That is the balance sheet’s first line of defense, and it is structural.

Return on invested capital47%
▲ uptrend · actual vs trend moving average · What the business earns on the capital tied up inside it.

Return on invested capital reached 47% in fiscal 2026 — against 10% a decade ago. Even the downcycle floor (28% in fiscal 2024) would flatter most industrial companies’ best years.

Capital returned to shareholders$4.9B
▲ uptrend · actual vs trend moving average · Cash handed back through buybacks and dividends.

Capital returns: $4.9B in fiscal 2026 — $1.3B of dividends (raised to $0.26 a quarter with the October 2025 release, accession 0000707549-25-000082) plus $3.7B of buybacks under the $10.0B authorization. The dividend has grown every year of the decade, downcycles included.

actual risingactual fallingtrend lineabove medianbelow median

How it adds up — the fortress test

Putting it together. An Altman Z-score of 34 — one of the highest readings the metric produces for a large industrial — a current ratio of 2.6, and $4.0B of long-term debt against $4.9B of annual free cash flow: under a year of cash generation covers every borrowed dollar. Deferred revenue of $2.4B (accession 0000707549-26-000033) means a meaningful slice of the coming quarters is already paid for. The balance sheet passed its most recent storm test — fiscal 2024 — with the dividend rising and research spending intact at $2.4B a year. On this tab’s question there is no debate to have. The forward cushion is unusual too: $2.4B of deferred revenue — customer cash already collected for tools and services not yet delivered — sits on the balance sheet as a paid-in-advance slice of the coming quarters (accession 0000707549-26-000033).

Your Rating
Is the balance sheet strong?
Could it survive a severe downturn?
Can it fund the growth from its own cash?
Notes
Next
Valuation →
What you’re being asked to pay for it today.
What’s Changed
The question — did this quarter strengthen or weaken the thesis?

The thesis was set when you decided to own the company; the only question now is whether the latest quarter made it stronger or weaker. Most of what crosses the tape is noise — a penny beat, a soft week, a worrying headline. This tab strips that away and asks what actually moved: did earnings and guidance confirm the story, or crack it? Did the multiple do something the business didn’t? Read it as a standing check on your reasons for owning the company. If those reasons are intact, the day-to-day price is somebody else’s problem.

Quarter ended Jun 2026 — the latest read · latest verified data
EPS, last 12 months
$5.76
+39% vs a year ago · a record fiscal year
P/E vs 5-yr mean
54.5×
+108% vs its ~26× norm — richer
Price · trend
$314
about 11% above its rising 1-yr average
The June quarter closed the best fiscal year in Lam’s history: $6.7B of revenue in the quarter (+30% on the year-ago quarter), $23.2B for the year, earnings above the top of the guided range for a fourth quarter in five, and September guided to $8.1B — one-fifth above the record just set (accession 0000707549-26-000033). The fiscal-2026 annual report followed on August 7 (accession 0000707549-26-000037), disclosing the numbers this report leans on: 36% of revenue from the installed base, foundry 54% and memory 39% of systems — and China at 33.8% of total revenue, the figure the Risk tab is built around. What changed for the stock was the price: a four-fold run to $433 by late June, then -28% back to $314. This is the first GARPify v6 read on Lam — it replaces an earlier draft page whose data was partly estimated; every figure here traces to the verified battery or an accession. The lights: quality green, wave green, price red, trend green-on-probation.
Guidance
With the results, management guided the September quarter to $8.1B ± $400M of revenue — one-fifth above the record just delivered — with $2.15 ± $0.15 of adjusted earnings per share (one-time items set aside) and a 52% ± 1% gross-margin guide, with $2.4B of deferred revenue already contracted toward it (accession 0000707549-26-000033). One quarter at a time, in explicit ranges, as always.
Thesis check — intact, with a price asterisk
The quarter strengthened every business light: a record year on every line, a fourth above-range earnings beat in five quarters, gross margin holding above fifty percent, and the boldest guide in company history sitting on contracted backlog. What it did not fix is the red light: at 54× versus a 26× five-year norm — the widest premium-to-history in this coverage — the June melt-up has only partly unwound. If you hold: nothing here argues the business is slipping. If you are waiting: the record quarter is already in the ticket price, twice over.
The last two years, weekly
$3142024-08-302026-08-2152-wk high $433
A History of What’s Changed

The recent year quarter by quarter (the cadence we pull), then a year-by-year look further back — each read straight from the verified figures: what the business did and what the market did. Each card carries the three lights for that period: Price and Trend computed from the period’s own data, Quality held at the report’s standing verdict (it is a judgment, not a number that flips each quarter).

The trailing year — quarter by quarter
2026-06 · Growth steady
QualityPriceTrend
12-month EPS $5.76 (+39%)P/E 75.2× (+187% vs 20-q norm)Price $433
The record close: $6.72B delivered against a $6.60B guide, EPS above the range again — and the quarter of the June price peak at $433, with a 78× multiple print at quarter-end that was the price spike, top-ticked.
2026-03 · Growth steady
QualityPriceTrend
12-month EPS $5.30 (+48%)P/E 40.3× (+71% vs 20-q norm)Price $214
A third straight above-range earnings quarter ($1.47 vs a $1.35±0.10 guide) as high-bandwidth-memory orders accelerated; the stock broke out through $214.
2025-12 · Growth steady
QualityPriceTrend
12-month EPS $4.88 (+48%)P/E 35.1× (+53% vs 20-q norm)Price $171
Above the range again ($1.27 vs $1.15±0.10); the December quarter confirmed the memory recovery was real, and the multiple pushed into the high thirties.
2025-09 · Growth steady
QualityPriceTrend
12-month EPS $4.54 (+47%)P/E 29.5× (+32% vs 20-q norm)Price $134
The fiscal year opened with a hit above the midpoint and a dividend raise ($0.23 → $0.26, accession 0000707549-25-000082); the re-rating was underway.
Further back — year by year (fiscal years, ended June)
FY2025
Rev +24%EPS +43%Op margin 32%Return on money invested 38%Yr-end P/E 23×Price (yr) -9%
The recovery year: revenue rebounded 24% and earnings 43% as memory spending returned and high-bandwidth memory turned Lam back into a growth story. The significance was in the mix: the recovery arrived foundry-first, beginning the share shift that defines the current revenue base.
FY2024
Rev -14%EPS -13%Op margin 29%Return on money invested 28%Yr-end P/E 37×Price (yr) +66%
The downcycle year, in full: revenue fell 14% and earnings 13% as the NAND bust bottomed — while free-cash-flow margin stayed near its highs and the dividend rose. Note the market’s sequence: the multiple collapsed to single digits the year BEFORE the earnings trough and was already re-rating as the numbers bottomed — the stock leads this business, in both directions, every cycle in this ledger.
FY2023
Rev +1%EPS +1%Op margin 30%Return on money invested 36%Yr-end P/E 19×Price (yr) +51%
Flat at the top: a record first half as the shortage boom crested, then memory orders stopped almost overnight. Revenue held; the order book and the multiple did not.
FY2022
Rev +17%EPS +22%Op margin 31%Return on money invested 41%Yr-end P/E 13×Price (yr) -35%
The supercycle: +17% on top of the prior year’s surge, with every fab and memory maker expanding at once — the boom whose unwind defined the next two years. The multiple peaked early and spent the year compressing even as earnings grew: the market smelled the digestion before the income statement did.
FY2021
Rev +46%EPS +78%Op margin 31%Return on money invested 35%Yr-end P/E 24×Price (yr) +101%
The breakout year: +46% revenue and +78% earnings as the pandemic chip shortage met the first wave of 3-D NAND scaling — the fastest growth in Lam’s modern history, and the year the market began re-rating the franchise from commodity cyclical toward structural grower.
FY2020
Rev +4%EPS +10%Op margin 27%Return on money invested 22%Yr-end P/E 21×Price (yr) +72%
The pandemic year: fabs were essential and kept building — revenue up 4% through the chaos, and the platform set for the two-year boom that followed.
FY2019
Rev -13%EPS +4%Op margin 26%Return on money invested 24%Yr-end P/E 14×Price (yr) +9%
The first downcycle in this ledger: revenue fell 13% as memory spending paused — and free-cash-flow margin stayed north of a quarter of revenue, the installed base doing quietly what it would do again in fiscal 2024.
FY2018
Rev +38%EPS +43%Op margin 29%Return on money invested 25%Yr-end P/E 13×Price (yr) +22%
The last year of the old normal: +38% revenue as 3-D NAND scaling hit its stride — the first demonstration of the layer-count arithmetic that powers the current boom.
This is the first GARPify v6 report on Lam, so the ledger above is built from the verified market and filing data rather than from prior editions; from the next quarterly refresh onward, each card records what changed against the previous read.
Your Rating
Did this quarter strengthen or weaken the thesis?
Notes
Feel free to score this after you’ve explored the tabs.
Next
Leadership →
Who is running it and whether their incentives align with yours.
Earnings Growth
STRONG
You’ve met who is driving the bus. Now open the hood.

GARPify rests on one idea, and it is Benjamin Graham’s: “In the short run the market is a voting machine, but in the long run it is a weighing machine.” Day to day, a share price is a vote — a show of hands driven by mood, the very sentiment you watched open a gap on the last tab. But over years the votes cancel out and the scale takes over, and what the scale weighs is earnings. Given enough time, a stock is pulled toward the profits underneath it.

That is why we spend so little time on the tape and so much on the engine. If earnings are the weight, the only questions that matter are how much this company can produce and whether the amount keeps growing — and earnings are built just three ways: more revenue × wider margins × fewer shares. That is the engine. Here is LRCX’s.

Is the company growing earnings consistently?
Powerfully, in waves: earnings per share compounded 27% a year over the decade to $5.76 — through two revenue recessions (fiscal 2019 and 2024), each followed by a recovery to new records within two years.
Is growth accelerating or decelerating?
Real fuel throughout: revenue roughly quadrupled over the decade to $23.2B, operating margins reached a record 35%, and the installed-base franchise (36% of revenue) made each downcycle shallower for cash flow than for orders.
Does management's guidance prove accurate?
The cycle is the caveat, in both directions: earnings swing harder than revenue, the five-year growth rate (16%) starts from a boom base, and fiscal 2024 cut EPS 13% before the current boom doubled it. Average across the cycle, not from trough to peak.
The earnings engine
What is driving the earnings — and which lever the EPS comes from
Product lines by S-curve · US operations, latest year, in the company’s own reported numbers
Systems (new tools)
64%
$14.9B
cyclical
Customer support & other (installed base)
36%
$8.3B
recurring
— system revenue: foundry*
54%
of systems
cyclical-growth
— system revenue: memory*
39%
of systems
deep-cyclical
Fiscal 2026 revenue of $23.2B (10-K accession 0000707549-26-000037). *Foundry/memory shares are of SYSTEM revenue only (logic/IDM is the remaining 7%); the two rows above them are the full-revenue split. Geography note: China was 33.8% of fiscal-2026 revenue — see the Risk tab.

Read it like a price chart. The actual metric (solid line, green when rising / red when falling) is plotted with its moving-average trend line (grey dashed) on the same axis — the fundamental version of a price and its 200-day average. The tag (uptrend / downtrend / range-bound) reads where the actual sits relative to that trend line. On mean-reverting metrics like margins, a tinted band also marks the level versus the metric's own 10-year median; on cumulative-growth metrics like EPS and revenue that band carries no signal, so it's omitted.

Revenue$23.2B
▲ uptrend · actual vs trend moving average · Real customer demand driving the top line.

Revenue reached $23.2B in fiscal 2026, up 26% — but read the whole line, because it is the most honest picture of this company’s character: down 13% in fiscal 2019, down 14% in fiscal 2024, and roughly quadrupled over the decade anyway. This is what cyclical growth looks like: the trend is powerful and the path is violent.

Earnings per share (TTM)$5.76
▲ uptrend · actual vs 1-yr moving average · The bottom line — what each share earns.

Earnings per share reached $5.76 in fiscal 2026 (split-adjusted), compounding 27% a year over the decade — and swinging harder than revenue in every cycle: the fiscal-2024 dip took EPS from $3.32 to $2.90 before the recovery more than doubled it in two years. Operating leverage cuts both ways; over any full cycle it has cut Lam’s way.

Operating margin35%
▲ uptrend · actual vs trend moving average · Profit kept on each dollar of sales.

Operating margin reached 35% in fiscal 2026 — a record — and gross margin crossed 50%, helped by mix and by the growing weight of installed-base services. A decade ago the operating line ran at 19%. Pricing power plus scale, demonstrated across two downcycles.

Shares outstanding1.25B
▼ shrinking · actual vs trend moving average · The buyback lever — flat or rising means EPS isn't being manufactured.

The share count fell from 1.60B to 1.25B over the decade — -22% — on steady buybacks ($3.7B in fiscal 2026 alone) under a $10.0B authorization (per the fiscal-2026 annual report, accession 0000707549-26-000037). Fewer shares, same machine: a seventh of the decade’s per-share growth came from this line.

actual risingactual fallingtrend lineabove medianbelow median

How it adds up — and what it's done to the price

Putting it together. The four lines describe a violent compounding machine: revenue quadrupling over a decade through two drawdowns, record margins, and a shrinking share count — earnings per share up 27% a year for ten years. Then the second term of the equation: over the past year the price rose 223% while earnings rose 39% — the remaining 133 points were multiple expansion. Over five years: price +383%, earnings +114%, multiple +125%. Roughly half the five-year return — and most of the last year’s — was the market raising the price per dollar of earnings. The engine is real and running at record output; the ticket now includes a large tip. One more decomposition worth carrying: of the decade’s roughly eleven-fold earnings rise, about a fifth came from the shrinking share count and the rest from operations — a business-growth story with buybacks as amplifier. The Leadership tab’s guide-and-deliver ledger draws the same picture over a shorter window.

Your Rating
Are earnings growing consistently?
Is the growth high quality, not manufactured?
Does management deliver on its guidance?
Notes
Next
Financial Strength →
Whether the balance sheet can survive a downturn and still fund the plan.
Sentiment
MIXED
In the short run the market votes. What is it voting for here?

This is the weighing machine’s noisy twin — the voting machine, live. Day to day, price is set by mood: how much the crowd likes the story right now. That mood is the gap you watched open between earnings and price, and it swings both ways. When a stock is adored, plenty of good news is already in the price and any stumble is punished; when it is ignored, expectations are low and the surprises tend to cut the other way.

So the question here is not whether LRCX is a good business — that is what the earlier tabs and your own read are for — but how much optimism the crowd has already paid for. A wonderful business everyone already loves is a very different bet from a wonderful business the market has overlooked. This tab reads where sentiment sits today, and how much room that leaves.

How does the market feel about this company right now?
Chastened-bullish, from a higher altitude: the market paid 78× at the June peak, took back 28% in eight weeks, and still pays +108% over the historical toll today.
Has sentiment shifted recently?
Re-rating did most of the work: of the past year’s 223% rise, earnings explain 39 points; the rest is the multiple. What the mood grants, the mood repossesses — July was the live demonstration.
Is the crowd right or wrong historically?
Both recent extremes were wrong within a year — the 9× despair of the memory bust and the 78× euphoria of June. The market’s record of pricing this franchise is poor in both directions; use its moods, do not adopt them.

"In the short run, the market is a voting machine; in the long run, it is a weighing machine." Graham’s line is this tab’s whole framework. The votes move the multiple; the scale weighs the earnings. Lam’s voting record is the most volatile in this coverage — which is precisely why it repays study.

A decade of voting: the same franchise priced anywhere from 9× to 78× per dollar of earnings. Both extremes came in the last four years, and both were repudiated within a year — the single-digit despair of the memory bust by a doubling of earnings, the 78× euphoria of June 2026 by a 28% drawdown in eight weeks.

Today’s 54× is +108% above the five-year average with the rolling-average line still being dragged upward by the spike itself. Against its theme-mate the comparison is instructive: similar franchises, similar waves — and Lam carries the larger premium to its own history, on the more cyclical earnings stream.

The decomposition quantifies the mood: of the past year’s 223% price rise, earnings explain 39 points; the remaining 133 points are re-rating. Over five years the multiple contributed +125% of its own. Re-rating is borrowed return — borrowed from future holders, at an interest rate set by the market’s patience.

A History of Lam's Sentiment Periods

What to watch is unchanged: the multiple against the earnings. The company will report record quarters — the $8.1B guide practically promises one. The stock’s return from $314 depends on whether the market keeps paying 54× for them. In fiscal 2024 it decided, abruptly, that it would not pay even 26×; in June 2026 it briefly paid 78×. Neither mood consulted the chambers.

2015-2018 - Cheap cyclical (PE ~19)
For most of the mid-2010s the market treated Lam as a commodity cyclical — low-to-mid-teens multiples, 19× at the end of 2016 — pricing every boom as if the next bust were already scheduled. It usually was; the stock compounded anyway.
2019-2020 - Grudging re-rating (PE ~22)
The fiscal-2019 downturn came and went without breaking the franchise, and the multiple crept to about 22× — the market beginning to concede that installed-base services had made the cash flows sturdier than the order book.
2021-2023 - Boom to despair (PE 22 → 9)
The shortage boom carried the multiple to about 22×; the memory bust that followed crushed it to 9× — single digits, the five-year low, the market pricing a permanent winter in NAND. It was the buying opportunity of the era. The range lesson stands: 9× to 78× for the same company, four years apart.
2024-2025 - The AI bid (PE ~22)
High-bandwidth memory turned Lam from a NAND-cycle story into an artificial-intelligence story, and the multiple settled around 22× at the end of 2024 — a premium to history, tethered to visibly accelerating earnings.
2025-2026 - Vertical, then air pocket (PE 78 → 54, today)
Then the tether snapped: a four-fold price run into the June 2026 peak put the multiple at 78× — territory previously reserved for earnings troughs, printed at an earnings record. The July air pocket took back 28%; at 54× today, sentiment remains +108% above the historical toll. Chastened, not chastened enough to be average.

One structural note: roughly 36% of revenue is now installed-base services (accession 0000707549-26-000037) — steadier than tool sales, and the best argument that Lam’s “normal” multiple deserves to sit above its own history. The argument has limits: services did not stop the fiscal-2024 revenue decline, and they will not stop the next one. They soften the cycle; they do not repeal it.

Stock Price — How the Market Has Voted Over Time
Earnings Per Share — What Sat Underneath the Price
Comparing the two reveals when sentiment (price) ran ahead of or behind fundamentals (earnings).
Your Rating
Is sentiment constructive rather than euphoric?
Are expectations modest enough to leave upside?
Is the stock under-appreciated for its quality?
Notes
Next
Technical →
Whether now is a stretched or a sensible moment to buy.
Valuation
RICH
If you’ve found a wonderful business, that’s only half the trade.

Everything up to here has been about the bus — the driver, the engine, the frame, the machine itself. But however good the machine, a business bought at any price is not automatically a good investment. This is where the discipline bites: you still have to buy it at a reasonable price. A great company and a great investment are not the same thing, and the difference between them is exactly what you pay.

Remember the weighing machine. Over time a price is dragged toward earnings — but the multiple you pay at the door decides how much of that growth lands in your pocket rather than the seller’s. Overpay for even the finest engine and the market can spend years merely growing back into your price. So this tab asks the plainest question in investing: at today’s price, is LRCX cheap, fair, or dear?

What are you paying for this business today?
The widest premium-to-history in this coverage: 54× trailing reported earnings against a 26× five-year average (range 9–78×), a PEG of 3.3 on the delivered 16% pace, and roughly 5 years of that pace prepaid before the multiple normalizes.
Is that cheap or expensive relative to history?
At the average multiple, current earnings price the stock near $151 — against $314 today. The record year and the $8.1B September guide are genuine; at 2.1 times the normal toll, they are also genuinely already paid for.
What does the market expect from here?
The earnings yield is 1.84% against a 4.707% Treasury — a negative 2.9-point spread on the most cyclical earnings stream in this coverage. Growth must not merely arrive; it must never pause.

"Price is what you pay; value is what you get." Buffett’s most-quoted line — inherited from Benjamin Graham — is the whole of the valuation discipline in seven words. A wonderful business bought at a foolish price is a poor investment. The prior tabs establish that Lam is a wonderful business; this one asks, more pointedly than anywhere else in this coverage, what today’s price assumes.

The anchors: over five years the daily multiple on reported earnings averaged 26.24×, bottomed at 9.03× in the 2022 downcycle despair, and peaked at 77.56× at the June 2026 price top. Today’s 54.5× sits +108% above the average — 2.1 times the toll the market has typically charged for a dollar of Lam’s earnings, the widest premium-to-history in this coverage.

The fair hearing: fiscal 2026 was a record on every line, the September guide of $8.1B ± $400M implies another enormous step, deferred revenue of $2.4B is already contracted, and the layer-count physics behind high-bandwidth memory is a genuine structural change (accessions 0000707549-26-000033, 0000707549-26-000037). If the artificial-intelligence build-out has permanently lifted Lam’s growth rate, the historical average understates fair value. That word — permanently — is carrying the entire ticket price.

The arithmetic without the narrative: at 54.5× the earnings yield is 1.84% against a 4.707% ten-year Treasury, a negative 2.9-point spread. We quote no forecasts — the price itself is the forecast: at the delivered 16%-a-year pace, earnings need about 5 more years of compounding before today’s multiple sits at its historical average. The PEG on five-year growth reads 3.3 against the GARP threshold of 1. And the five-year average multiple prices current earnings near $151 — less than half of today’s quote. Every anchor pulls the same direction.

Verdict: the reddest reading on the panel, at a company whose own history argues both sides. The chart below maps Lam against the other tollbooths of the build-out on multiple versus growth — the whole neighborhood is expensive, and Lam is priced at the aggressive end of it relative to its own past. The discipline is unchanged: say plainly that the business is superb, and that at $314 the price already contains a decade in which nothing cyclical happens to a deeply cyclical company.

PE Ratio vs Its 5-Year Average
The multiple today sits far above its 5-year average of 26.24× (daily range 9.03×–77.56×, the top set at the June 2026 price peak) — you are paying roughly 2.1× what the market has typically charged per dollar of these earnings.
Valuation vs Growth — LRCX and Its Basket Peers
PE ratio (left) against 5-year EPS growth (right) for LRCX and peers. Low multiple paired with high growth is the GARP signature.
Your Rating
Is the price reasonable for the quality?
Is it cheap versus its own history?
Does the growth justify the multiple?
Notes
Next
Risk →
The ways the thesis could break — and how survivable each is.
Leadership
STRONG
Who’s driving the bus?

GARPify focuses on strong management, and we judge it by one thing: the work. A company’s future earnings come from what its leaders actually do — the plan they set, and their record of keeping it. A share price is that same work plus one thing the team does not control: the market’s mood.

So this tab reads the work first — who runs the business, what they have built, what they intend next, and whether they keep their word — and only at the very end lays it against the price. The gap between the two is where this tab is headed.

Who is running this company?
Long-tenured engineers with a clean ledger: five quarters guided since mid-2025, five delivered at or above the midpoint, four above the entire guided range — and one orderly, internally-promoted operating transition (accession 0000707549-26-000014).
How long have they been here and what is their track record?
Capital allocation is systematic: a dividend raised every year of the decade, $3.9B of fiscal-2026 buybacks under a $10.0B authorization, and a share count down 22% in ten years.
Do they have skin in the game?
The guidance culture is the tell: one quarter at a time, explicit ranges, systematically conservative — the September guide of $8.1B is bold precisely because this team does not guide casually.

Jim Collins, in studying what separates enduring great companies from merely good ones, kept returning to a quiet variable: the quality and continuity of the people at the top. Not charisma — continuity. The companies that compounded for decades tended to be run by people who thought in decades. Lam’s leadership reads straight from that playbook.

The chief executive has run the company since late 2018; the chief financial officer’s signature has anchored the filings for years alongside him (fiscal-2026 annual report, accession 0000707549-26-000037). The one senior change in two years was handled the way good engineering companies handle succession: the chief operating officer’s retirement announced with a sitting internal successor, effective on a stated date (accession 0000707549-26-000014). No drama, no search firms, no gap.

Capital allocation runs on the same rails every year: fund the research first ($2.4B in fiscal 2026), raise the dividend (every year of the decade, now $0.26 a quarter), retire shares with what remains ($3.9B in fiscal 2026, 22% of the count over ten years). Repeatable decisions, repeated.

The governance picture is mature and technical: deep engineering leadership, orderly internal succession, and a guidance culture of explicit quarterly ranges that the company then meets or beats. This is leadership as stewardship of a franchise built on physics — execution rather than vision as the watch-word.

What management is saying now

Listen to what this company promises and you hear the same modest sentence every ninety days: here is next quarter, in a range, and here is how we did against the last one. No five-year visions. From a team that keeps landing above the top of its own guidance, the restraint is the message.

Highlights of most recent guidance · with Q4 FY2026 results, July 29, 2026 (accession 0000707549-26-000033)

Sept-quarter revenue
$8.1B ± $400M
a record by ~20% if delivered
Sept-quarter EPS (adjusted*)
$2.15 ± $0.15
vs $1.82 just delivered
Gross margin guide
52% ± 1%
adjusted*
Dividend
$0.26/quarter
raised from $0.23 in Oct 2025

The September guide is the boldest promise in this coverage — a 21% sequential step to a quarter one-fifth larger than the record just set. Management guides one quarter at a time inside explicit ranges; the ledger below is why the market takes the ranges seriously.

Ask this management what the business is, and the answer is the same every quarter — which is the point. The durable message has four parts. First, layer physics: every chip generation needs more deposition and more etch, so Lam’s served market grows faster than wafer starts — the structural argument under the cyclical noise. Second, the installed base: every chamber shipped becomes an annuity, and the fleet only grows. Third, cycle honesty: management neither denies the cycle nor apologizes for it — it guides one quarter at a time and lets the ranges carry the message. Fourth, capital return: a dividend raised every year for a decade and a $10.0B authorization behind the buyback — the cash comes back, boom or bust (accessions 0000707549-25-000082, 0000707549-26-000037).

How much should you trust what they say? Quantifiably: across the five graded quarters in the Leadership ledger, delivery averaged +2.6% versus guided revenue midpoints and +9.1% versus guided earnings midpoints — zero range misses, four earnings prints above the entire range. When habitually conservative speakers make the boldest promise in company history, the base rate says treat the midpoint as a floor — and watch it like the single most informative number of the quarter, because it is. The number after that: the gross-margin print against the guided corridor, the tell on whether record volume is coming at record quality.

The record
One quarter at a time. Guidance is offered a single quarter ahead, inside explicit ranges — an admission that cyclical visibility is short, and a discipline the record rewards.
Said and then beaten. The latest guide: about $8.1B of revenue and $2.15 of adjusted earnings per share for the September quarter — the boldest in company history, from a team that keeps clearing its own ranges.

The business they run

Every leading chip in the world is built layer by layer — etched away and deposited back, thousands of times — and two companies make the machines that do it. Lam is one of them. When the fabs of the world expand, they cannot route around it; when they pause, the installed base keeps paying. One franchise, 50-cent gross margins, and a service annuity that bills through every winter.

Lam Research builds the machines that sculpt chips: plasma-etch systems that carve features finer than a wavelength of light, and deposition systems that lay films down one atomic layer at a time. Every leading chipmaker runs its chambers. The useful way to see Lam is as a portfolio of S-curves — separate demand curves at different points in their lives: one steepening curve (foundry systems, pulled by the artificial-intelligence build-out), one violent cyclical (memory systems, today booming on high-bandwidth memory), one fading side-curve, and under all of them a recurring installed-base annuity that grows through every winter. In fiscal 2026, revenue of $23.2B split four ways:

Systems — foundry
New etch and deposition tools shipped to the contract chipmakers building advanced logic — the artificial-intelligence build-out’s front line, where gate-all-around transistors and advanced packaging both add etch and deposition steps. The share of systems revenue went from 38% to 54% in three years and the dollars nearly doubled — the steepening curve that made fiscal 2026 a record.
Percentage of Sales35%
3-Year Sales Growth Rate+26% / yr
Slope of CurveSTEEP
Systems — memory
Tools for DRAM and 3-D NAND makers — the most violently cyclical customers in electronics, and the reason Lam’s history has 40%-plus revenue swings. High-bandwidth memory is today’s pull; the 2022–23 bust is the base rate.
Percentage of Sales25%
3-Year Sales Growth Rate+9% / yr
Slope of CurveCYCLICAL
Systems — logic & other
Trailing-edge and specialty system sales. The shrinking share is mostly arithmetic — foundry grew past it — but the dollars also fell as customers concentrated spending at the leading edge.
Percentage of Sales4%
3-Year Sales Growth Rate-21% / yr
Slope of CurveCYCLICAL
Customer support & other (installed base)
Spares, service and upgrades on every chamber Lam has ever installed — a fleet that only grows, billing regardless of the order book. This line grew straight through the fiscal-2024 downcycle and keeps free cash flow near its highs when tool orders pause: the annuity under the cycle, and the quiet argument for a higher-than-historical multiple.
Percentage of Sales36%
3-Year Sales Growth Rate+7% / yr
Slope of CurveRECURRING
Shares are fiscal-2026 revenue: the systems/customer-support split from the FY2026 Form 10-K (accession 0000707549-26-000037), with systems allocated across markets by the 10-K’s disclosed percentages (foundry 54%, memory 39%, logic & other 7% of systems). Growth is the 3-year annualized rate FY2023–FY2026 on the same basis (FY2023 from the FY2024 Form 10-K, accession 0000707549-24-000106); market-line growth is approximate because it applies disclosed percentages to disclosed totals.

The model has the same three gears as the rest of the theme. Sell the tool: $14.9B of systems revenue in fiscal 2026. Service the fleet: $8.3B from the installed base — the 36% annuity above. Ride the physics: every chip generation needs more layers, and every layer is etched and deposited. The result: 50% gross margins, 35% operating margins, a 49% return on invested capital, and $4.9B of free cash flow — monopoly-class economics from machines that weigh tons.

Two things to hold from this page as you read on. First, the mix explains the volatility: nearly two-thirds of revenue is new-tool sales into the most cyclical buying decision in electronics, which is why Lam’s history includes 40%-plus revenue swings that its income statement then survives in style. Second, the annuity explains the survival: an installed-base business compounding 7% a year that kept growing straight through the fiscal-2024 downcycle is what separates a great cyclical from a great business that happens to be cyclical. The Financial Strength tab shows what that annuity does to the cash flows; the Risk tab shows the two concentrations — China and memory — that ride along with the franchise; and the Valuation tab shows what the market now charges for the whole package.

Read the mix shift for what it says about the cycle: foundry went from 38% of systems revenue in fiscal 2023 to 54% in fiscal 2026 while logic’s share fell from 20% to 7% — the artificial-intelligence build-out has concentrated Lam’s tool sales onto the leading edge at remarkable speed. Memory’s share held near 39% throughout, but its composition changed underneath: high-bandwidth memory for AI accelerators is pulling the same etch-and-deposition intensity that 3-D NAND scaling did in the last cycle. Concentration cuts both ways here exactly as it does at its theme-mate — it is why the margins expanded to records, and why the Risk tab ranks the memory cycle and China among the high risks: there is no uncorrelated second business to hide in.

And read the annuity for what it says about the future: at 36% of revenue, customer support grew +7% / yr through a period that included a full downcycle — every chamber shipped in this boom becomes service revenue in the next winter, so the installed base is the one line the cycle permanently, cumulatively enlarges. What would change the slopes on the cards above: watch layer counts (they only rise), the China share (falling — from about 42% of revenue two fiscal years ago to 33.8% now, per the 10-Ks), and the ratio of service growth to systems growth, the quiet indicator of how much of each boom is being banked into the annuity.

Two mechanics worth understanding before the tabs ahead. The upgrade economics: a fab rarely rips out a Lam chamber — it converts it, buying upgrade kits and process retrofits through the service line to push an installed tool to the next node, which is why the customer-support business behaves like a subscription on the world’s fab capacity rather than a spares counter. And the forward cushion: $2.4B of deferred revenue — customer cash already collected for undelivered tools and services — sat on the balance sheet at year-end (accession 0000707549-26-000033), covering roughly 30% of the September quarter’s guided midpoint before a single new order. Competition at the leading edge is real but short-listed — a handful of names, each defending decades of installed process libraries — and the 50% gross margin is the market’s running verdict on how that contest stands.

The record
Two indispensable names. Etch and deposition at the leading edge is a two-and-a-half-horse race behind decades of installed process libraries — and Lam holds or gains position through every node transition.
The annuity is the shock absorber. Installed-base revenue — 36% of the total — kept free-cash-flow margin near 29 cents on the dollar straight through the fiscal-2024 downcycle.
Software economics in a hard hat. 50% gross margins, 35% operating margins, 49% return on the money invested, $4.9B of annual free cash flow.

The team running the business

The same engineers have run this company for years, and the plan they run barely changes from one node to the next. In an industry that reinvents itself every eighteen months, the strategy is deliberately the boring part — and the one senior change in two years was handled the way good engineering firms handle succession.

An engineering company run by insiders with long tenures — and one recent, orderly transition at the operating level: the chief operating officer’s retirement and an internal promotion to succeed him, disclosed in February 2026 (accession 0000707549-26-000014). The top two signatures on the filings have not changed.

Timothy M. Archer
President and Chief Executive Officer
Chief executive since late 2018 and a director alongside (per the fiscal-2026 annual report, accession 0000707549-26-000037). His tenure spans the fiscal-2019 downturn, the pandemic, the memory bust and the artificial-intelligence build-out — through which revenue roughly doubled and the installed-base franchise became a third of the company.
Douglas R. Bettinger
Executive Vice President and Chief Financial Officer
Keeper of the guide-and-deliver ledger below — five quarters guided, five landed, four of them with earnings above the top of the guided range — and of a capital-return machine that paid $1.3B in dividends and repurchased $3.9B of stock in fiscal 2026 (accession 0000707549-26-000033).
The operating bench
COO transition, March 2026
The one senior change in two years was a model succession: the chief operating officer’s retirement announced with a sitting internal successor already named and a stated effective date (8-K accession 0000707549-26-000014). Depth promoted from within, disclosed on time — governance working the way the filings say it should.

A note on how this report judges a team it deliberately profiles thinly: GARPify cites only what filings confirm — titles, signatures, disclosed actions — and skips the interview-circuit color. By that standard the evidence here is unusually strong: stable signatures across years of filings, one succession handled with an internal promotion and a stated date, a dividend raised every year for a decade including through a 14% revenue decline, and a guidance ledger whose misses number zero. Management quality is an output you read in the ledger, not a personality you assess in a profile.

The record
Continuity at the top. The chief executive has run the company since late 2018; the finance chief’s signature has anchored the filings for years (accession 0000707549-26-000037).
Succession without drama. The one senior change in two years: the chief operating officer’s retirement announced with a sitting internal successor, effective on a stated date (accession 0000707549-26-000014).
Guidance is a promise they keep. Five straight quarters guided and landed, four with earnings above the entire guided range — the ledger is on the Accomplishments card.

What they’ve accomplished

Read one two-year stretch before you read the rest: fiscal 2024, when revenue fell 14% and earnings fell with it — and what the team did next. It raised the dividend, kept buying stock, protected the research budget, and grew the installed-base annuity. That is the whole case for trusting them through the next downturn, because there will be one.

A decade ago Lam was one strong equipment maker among several. Under this team it became one of the two names modern chipmaking cannot be built without — and it did so through the most violent demand swings of any company in this coverage. The résumé, fiscal 2016 to fiscal 2026 (per-share figures split-adjusted):

Revenue$5.9B→$23.2B
Op. margin19%→35%
Return on money invested10%→47%
Share count1.60B→1.25B (split-adjusted)
Diluted EPS$0.52→$5.76

Those numbers are the result. What produced them was four levers, pulled consistently:

1
Etch and deposition leadership, held through every transition.
As chips went three-dimensional, the hardest problems in the fab migrated toward exactly what Lam builds — and $2.4B a year of research kept it on the short list at every node. The 50% gross margin is a decade of that position, priced.
2
An installed-base annuity built chamber by chamber.
Customer-support revenue grew from $6.7B to $8.3B in three years (+7% / yr) — recurring revenue from every chamber ever installed, the reason cash flow barely noticed the fiscal-2024 downcycle.
3
Cycles navigated without franchise damage.
Two revenue recessions inside the decade — fiscal 2019 and fiscal 2024’s 14% drop — and through both: research funded, dividend raised, share count still falling. The proof is fiscal 2026: the record year arrived with the franchise fully intact.
4
Relentless, boring capital return.
The share count fell 22% over the decade; the dividend rose every single year; fiscal 2026 returned $5.1B under a $10.0B authorization (accessions 0000707549-26-000033, 0000707549-26-000037).

It matters which lever did the lifting: the first two — position and annuity — because they are the ones a downturn cannot repossess. The fiscal year just closed is the exhibit: $23.2B of record revenue, margins at all-time highs, and every quarter of it guided in advance with four of five delivered above the entire promised range.

The decade’s cash ledger makes the fourth lever concrete: fiscal 2017 through 2026, Lam generated $34.1B of free cash flow and returned $33.9B — $7.8B in dividends plus $26.2B in buybacks, roughly 100% of everything the business threw off (YCharts annual series). Across a full cycle, the revealed capital policy is: fund the research first, then hand back essentially all the rest. The stress exhibit is fiscal 2024: revenue down 14%, and the dividend rose anyway — the year to reread when the next downcycle headline arrives.

A structural note completes the résumé: over just the last three fiscal years the installed-base annuity grew from $6.7B to $8.3B while its share of revenue held near 36% of a much larger company — the team has been converting each boom into permanent recurring revenue, which is the single accomplishment most likely to matter in the next downturn.
The record
A decade of cyclical compounding. Revenue roughly quadrupled to $23.2B, earnings per share $0.52→$5.76, the share count down 22%.
The stress test they passed. Fiscal 2024 cut revenue 14% and earnings 13% — and the franchise, the dividend and the research budget all came through intact.
Cash handed back. $5.1B returned to owners in fiscal 2026 alone, behind a $10.0B buyback authorization.

The plan for the future

The plan is almost aggressively unglamorous: hold the one franchise that cannot be skipped, spend heavily to defend it, convert each boom into permanent installed-base revenue, and hand the rest of the cash back. Watch what they DO with the money — that is where the strategy actually lives.

A plan is really just the next promise — and Lam makes its promises quarterly: each earnings release guides the coming quarter’s revenue and adjusted earnings (one-time items set aside) inside explicit ± ranges. That cadence is the plan made visible: no investor-day theatrics — hold the etch-and-deposition position with research spending ({fmtB(E_RD)} in fiscal 2026), grow the installed-base annuity, return the surplus through a dividend raised every year and the {fmtB(BUYBACK_AUTH)} buyback authorization, and prove it one range at a time. Management guides exactly two numbers each quarter — revenue and adjusted earnings (one-time items set aside), both with explicit ± bands — plus a gross-margin corridor; nothing annual, nothing multi-year, nothing to quietly walk down later. Below, the ledger for the past five quarters, plus the promise currently on the table — all post the October 2024 ten-for-one split (accession 0000707549-24-000123), so every figure is on today’s share basis.

Do they do what they say?

QuarterGuided revenueDeliveredVerdictGuided EPS*DeliveredVerdict
Q4 FY25$5.00B ± $300M$5.17Bhit$1.20 ± $0.10$1.33beat — above the range
Q1 FY26$5.20B ± $300M$5.32Bhit$1.20 ± $0.10$1.26hit
Q2 FY26$5.20B ± $300M$5.34Bhit$1.15 ± $0.10$1.27beat — above the range
Q3 FY26$5.70B ± $300M$5.84Bhit$1.35 ± $0.10$1.47beat — above the range
Q4 FY26$6.60B ± $400M$6.72Bhit$1.65 ± $0.15$1.82beat — above the range
Q1 FY27$8.10B ± $400Min progress$2.15 ± $0.15reports late Oct
*Adjusted diluted EPS (one-time items set aside), the instrument management guides. Verdicts grade delivery against the guided range: inside = hit, above the top = beat. Sources: 8-K accessions 0000707549-25-000051, 0000707549-25-000068, 0000707549-25-000082, 0000707549-26-000006, 0000707549-26-000020, 0000707549-26-000033.

The habit worth watching in that table: revenue lands inside the range every time, at or above the midpoint — and earnings clears the ENTIRE range four times in five. That is not luck; it is a guidance culture calibrated to under-promise by roughly one buffer’s width, quarter after quarter. It reframes the September guide: a promise one-fifth above the record just delivered, made by a team whose promises have been floors, not targets. Either the demand is extraordinary, or the culture just changed. The ledger says bet on the former.

What the promises earned you

Three lines, each indexed to 1.0 at the Q4 FY25 quarter: what management guided, what it Delivered, and what the market paid for the shares. Guidance rose 1.38× and delivery tracked it at 1.37× — from a base quarter that was itself an above-range beat, with four of five quarters clearing the entire guided range. The market line, 4.5× over the same five quarters, is the multiple doing the rest — the Valuation tab’s subject.

Q4’25Q1’26Q2’26Q3’26Q4’26market 4.5×delivered 1.37×guided 1.38×

The plan from here, as guided: a $8.1B September quarter — one-fifth larger than the record just delivered — with $2.4B of deferred revenue already contracted toward it (accession 0000707549-26-000033). The ledger says take the range seriously. The Valuation tab notes the market already has.

The promise on the table
$8.1B ± $400M
Sept-quarter revenue — +20% sequential
Earnings promised with it
$2.15 ± $0.15
adjusted*
Already contracted
$2.4B
deferred revenue at year-end

What to watch from here, in order: whether the $8.1B promise lands (late October, the next graded row); whether the memory share of systems — 39% in fiscal 2026 — keeps its footing when high-bandwidth-memory ordering normalizes; and whether China’s 33.8% revenue share keeps drifting down as the rest of the world outgrows it. The plan is simple. The ledger is why it is credible.

Two computations put the ledger’s meaning in numbers. Across the five graded quarters, delivered revenue averaged +2.6% versus the guided midpoint and delivered earnings averaged +9.1% — the largest average earnings beat in this coverage, four times clearing the entire guided range. Apply that habit to September and the $8.1B midpoint reads as a floor with history behind it. And note the promised scale: $8.1B annualizes to $32.4B, roughly 39% above the record year just closed, before any growth through fiscal 2027. The 54× multiple is pricing that trajectory and more; the ledger says the trajectory is credible, and the Valuation tab prices the “more.”

A note on holding a quarterly guider: there is no full-year number to anchor on — the information arrives in twelve-week increments, and each can reset the story. That makes this ledger the report’s best early-warning instrument, with an asymmetry worth naming: at the widest premium-to-history in this coverage, a clean guidance record is mostly downside protection that vanishes the first time it breaks. The first quarter to land below its guided midpoint — unprecedented in this sample — would hit the earnings line and the multiple at once. Until then: defend the position, grow the annuity, return the cash, and say exactly what the next quarter will bring.

One more reading of the September promise, against the balance sheet rather than the ledger: $2.4B of deferred revenue means roughly 30% of the guided quarter is already collected cash awaiting delivery — the boldest guide in company history is, in meaningful part, describing orders that are already paid for. That is the strongest kind of guidance a cyclical can give, and it frames the gross-margin corridor guided alongside it (52% ± 1%): management is promising record volume WITHOUT mix degradation. If both land, fiscal 2027 opens at a run-rate the Valuation tab’s red light has already paid for; if either slips, the ledger records its first blemish at the worst possible multiple. Late October answers.

The record
Defend the franchise. $2.4B of research a year, protected through every downcycle — the moat is maintained on purpose, not by luck.
Convert the boom. Each cycle turns new-tool sales into recurring installed-base revenue — now 36% of the total, the line that made fiscal 2024 shallower for cash than for orders.
Return the rest. $5.1B back to owners in fiscal 2026; the share count has fallen 22% over the decade, behind a $10.0B authorization.
Your Rating
Is this a capable, disciplined team with a proven record?
Is the plan credible and the growth durable?
Do they do what they say, and can you trust the guidance?
Notes
Next
Earnings Engine →
Where the earnings come from, and why the engine — not the mood — does the work.
Risk
ELEVATED
Now the hard part: what could go wrong?

The earlier tabs laid out the business, the people, the balance sheet and the price — the evidence for the investment. This tab does the opposite, on purpose: the surest way to lose money is to settle on a view and stop hunting for the holes in it. So here we go looking for what could make the thesis fail.

Every danger is worth weighing on two axes: how likely it is, and whether the business survives it. A risk that costs a year of growth is a world apart from one that breaks the engine for good — and the balance sheet from the last tab is what usually decides which. We lay out the handful of ways it could go wrong; the weighing is yours.

"Risk comes from not knowing what you are doing." The point of this tab is to know exactly what you are doing: owning the most cyclical franchise in this coverage, with a third of its revenue inside an export-control regime and two-fifths of its systems sold to memory makers, at the widest premium to its own valuation history — in exchange for arguably the strongest demand tailwind any equipment maker has ever had. Every risk below is visible in the ten-year record; the discipline is to weigh them before taking the position, not after.

What could go wrong?
The ticket, first: 2.1 times the historical multiple means reversion alone — no business problem — prices the stock near $151. The July air pocket was the mechanism firing once, gently.
How likely is it?
Then the concentrations: China at 33.8% of revenue inside a hardening export-control regime, and memory at 39% of systems — the most cyclical customer set in electronics, currently in its biggest boom.
Has management acknowledged it?
The franchise itself is the safe part: a duopoly-grade position defended by $2.4B a year of research, record margins, and an installed base that pays through every winter.
The dangers at Lam are almost entirely outside the machines. The franchise is the safest thing in the package — a duopoly-grade position in etch and deposition, 50% gross margins, an Altman safety score of 34, and an installed base that bills through every winter. The danger lives in the ticket and the customer set: the richest multiple in this coverage, a third of revenue inside an export-control regime, and the most cyclical customers in electronics. The register below is graded to match — two elevated, one more, and the balance-sheet risk near zero. What each one is, how serious it is, and what to watch.
ELEVATEDThe multiple itself
54× reported earnings against a 26× five-year average is the widest premium-to-history in this coverage, bought at a 2.9-point negative spread to the risk-free rate. Reversion to the average multiple — with earnings unchanged — is a price near $151. July 2026 demonstrated the mechanism: 28% erased in eight weeks, no bad news required. Watch the price-to-earnings multiple against its own rolling average on the Sentiment tab, and any quarter where the price falls while earnings rise — that is the multiple deflating in real time. The table below sizes it exactly, with today’s earnings held flat:
If the market pays…MultipleImplied pricevs today
One more delivered year, today’s multiple holds54×$399+27%
Today — earnings flat, multiple holds54×$314+0%
De-rate to the 5-year average multiple26×$151-52%
De-rate to the 2022–23 downcycle-low multiple$52-83%
Prices = the stated multiple × last-twelve-months earnings of $5.76 (top row grows earnings one year at the delivered 27% ten-year pace). Not forecasts — arithmetic, to size the downside an owner is actually exposed to. Today’s price $314; 52-week range $97–$433.
ELEVATEDChina concentration
China was 33.8% of fiscal-2026 revenue — $7.9B (10-K accession 0000707549-26-000037) — the largest single geography, inside an export-control regime that has tightened repeatedly and can change without warning. The mitigant is arithmetic, not comfort: the other two-thirds of revenue grew fast enough to set records anyway, and the comparable figure was about 42% two fiscal years ago — the dependence is falling. It remains the single most concentrated policy exposure in this coverage. Watch the geography split at each 10-K and 10-Q, and any new export-control rule that names deposition or etch tools.
ELEVATEDThe memory cycle
39% of system revenue ships to memory makers — the most violently cyclical customers in electronics. The 2022–23 NAND bust took Lam’s revenue down 14% and its multiple to 9×. High-bandwidth memory is today’s boom and it is real; it is also memory, and memory has never yet skipped a bust. The shock absorber is the 36% of revenue from the installed base, which bills regardless of new-tool orders. Watch memory-maker capital-spending announcements, and the foundry-versus-memory mix at each 10-K.
MEDIUMCustomer and geographic concentration
Beyond China, the customer list is short by nature — a handful of giant chipmakers concentrated in East Asia. One deferred fab program moves a quarter; regional risk around Taiwan and Korea is the industry’s permanent weather. This is the shape of the business, priced in by sophisticated holders, but an owner should know the fortunes ride on few decision-makers. Watch the deferred-revenue and geography disclosures each quarter — deferred revenue currently $2.4B, accession 0000707549-26-000033.
LOWCompetition and technology transitions
Etch and deposition at the leading edge is effectively a two-and-a-half-horse race with enormous entry barriers — installed bases, process libraries, $2.4B a year of research — and Lam has held or gained position through every recent node transition. A 50% gross margin is the market’s testimony. The franchise risk is a discontinuity in how chips are built; none visible today routes around plasma. Watch the gross-margin trend, and any leading chipmaker qualifying a new supplier at the leading edge.
The shape of the register is the point. Read the scenario table’s bottom row as the floor to size for: a full round-trip to the 2022–23 downcycle-low multiple, earnings held flat, is a 83% drawdown — the kind of weather this stock has actually delivered inside a single year. The fair-value area near $151 sits closer, a 52% give-back. The top row is the friendlier reality: one more delivered year at the ten-year pace with the multiple merely holding, and the price is HIGHER. The register reads ELEVATED, not extreme, because the top three risks are the price, a policy line and a cycle — all external to the machines — and the franchise itself, defended by physics and $2.4B a year of research, sits at the bottom.
The record
Rank order. The multiple, then China, then the memory cycle — all external to the machines — then customer concentration, then a low competitive risk the margins already refute.
The one number to watch. The price-to-earnings multiple against its own 26× average. At 54× it prices roughly 5 years of the delivered pace before it normalizes; the July air pocket was that mechanism firing once, gently.
The floor. A round-trip to the downcycle-low multiple on flat earnings is about a 83% drawdown; the fair-value area near $151 is the nearer 52% give-back.
What we are NOT worried about. The balance sheet or the franchise — Altman 34, $4.9B of annual free cash flow against $4.0B of long-term debt, and a 50% gross margin. The machines and the money are the safe part.
Your Rating
Are the key risks understood and disclosed?
Are they survivable if they hit?
Is the risk acceptable for a 3-5 year holder?
Notes
Next
Sentiment →
How much optimism the crowd has already priced in.
Technical
STRONG
Whatever you make of the price, is now a sensible moment to climb aboard?

A chart cannot tell you what a business is worth — that is for the earlier tabs and your own judgment. What it can show is the terms the market is offering today: whether the stock is stretched far above its own trend and priced for perfection, or has cooled to a calmer, more sensible entry. This is timing in the mildest sense — not prediction, just noticing whether you are chasing the bus or being offered a reasonable seat.

It matters because even a fair price can be bought at an unfortunate moment, right after a sharp run. So this tab reads LRCX’s trend and how far price sits from it — the tactical complement to the valuation question, and the last check before conviction becomes a decision.

What phase is this stock in?
Constructive on the house rule: $314 sits about 11% above a 1-year average still rising — the widest cushion in this coverage, +11%.
Where is price relative to the 1-year moving average?
The slope flatters: the four-fold run to $433 is still inside the 30-week window, so the average’s +14% pace will fade mechanically as those weeks roll off — expect the cushion to narrow even at a flat price.
What does the volume pattern say?
The line: a weekly close decisively below the $284 area with the average flattening flips the light to broken — and at 54× earnings, broken trend plus richest-in-coverage multiple is the framework’s hardest stop.

The technical read in this report is deliberately primitive: one price, one 1-year moving average, and the direction of that average. No patterns, no oscillators, no targets. The single question is whether the market’s weighing machine is currently trending with you or against you.

The current read, stated fully: price $314, which is about 11% above the 1-year weekly average of $284; the average itself is rising, up +14.2% over the past eight weeks — the steepest average and the widest cushion in this coverage. The house rule grades exactly three states: above a rising average is constructive, below a falling one is broken, everything else is a hold-your-fire. By that rule this chart is constructive, and has been through essentially the entire post-2022 advance.

The same two footnotes apply, scaled up. The average’s steep slope is the June vertical still inside the 30-week window — it will flatten mechanically as those weeks age out. And the 28% July drawdown is recent trend damage: the market has demonstrated the speed at which it reprices this name, in both directions, within a single quarter.

The 52-week range is the emotional bill of ownership: $97 to $433 — a 347% span in one year, the widest in this coverage. Size the position for the demonstrated weather: this stock’s history includes forty-plus-percent drawdowns with the business intact.

A History of Lam's Technical Periods

What would change the read: a weekly close decisively below the $284 area with the average flattening flips this light to broken — and at 54× reported earnings, a broken trend here is the framework’s full-stop combination, not a dip to average into.

2015-2018 - The first march ($8 → $14)
Split-adjusted, from about $8 to $14 in four years — a strong cyclical uptrend, punctuated by the late-2018 downturn that closed the era with a sharp correction.
2019-2020 - Shock and V ($29 → $19 → on)
A powerful 2019 recovery ran straight into the COVID crash — about $19 at the March 2020 low — and out the other side nearly vertically. The house pattern: macro shocks in this name have been entries, because the earnings machine kept running.
2021-2022 - The memory bust (peak ~$73, trough ~$31)
From the shortage-boom peak near $73 the stock lost roughly half its value into late 2022 as memory spending collapsed. The base rate for position sizing: drawdowns of forty-to-fifty percent are part of this stock’s normal weather, franchise intact throughout.
2023-2025 - The AI trend ($31 → $72 → on)
From the 2022 trough the stock ran a textbook uptrend — higher highs, higher lows, price above a rising 1-year average through essentially all of 2023–2025 as high-bandwidth memory re-wrote the demand curve.
2025-2026 - Vertical and air pocket ($97 → $433 → $314, today)
The trend went parabolic: better than four-fold inside the year to the June peak at $433, then -28% in eight weeks. Today’s $314 holds +11% above a 1-year average still rising at +14% per eight weeks — a slope that is mostly the spike itself, still inside the window. Intact, stress-tested once, watched closely.

Overhead supply is the near-term texture: everyone who bought between $314 and $433 in the past three months is underwater, and rallies into the $390–$433 zone will meet their relief selling. Bases take time; patience is a position.

Weekly Close vs 1-Year Moving Average
Weekly closing price and its 1-year moving average. Price above a rising average indicates an intact uptrend.
Your Rating
Is the long-term trend intact?
Is the entry point reasonable, not stretched?
Is now a sensible time to buy?
Notes
Next
Think →
Where our work ends and your decision begins.
Think
Your Ratings — Updated as you go
Section Your Rating Notes
Rate each section as you read — your results appear here automatically.
GARPify did the work. Now you do the thinking.

Every number in this report is only useful if it changes how you make a decision. Most investors read, nod, and move on. The ones who build wealth over time do something different — they write things down.

Not because writing is the point. Because writing forces you to be honest with yourself. You cannot write "I believe this business will compound at 15% for the next decade" without immediately knowing whether you actually believe it.

These five questions have no right answers. GARPify has no opinion on what you should decide. That is your job. This is just the structure that makes the job easier.

Take ten minutes. Be honest.

Part 1
My Decision Process
Answer these five questions in your own words before you decide. Writing forces clarity. Clarity prevents mistakes.
1. What is this business actually doing that creates value?
In your own words — not GARPify's. If you can't explain it simply, you don't own it yet.
2. What price tells you the market is being irrational — and why do you believe that?
Not what the report says. What do YOU believe about fair value and why.
3. What has to go right for this to be a good investment over three years?
Name the two or three things you're betting on. Be specific.
4. What would make you sell — what fact or event would change your thesis?
Write this down now, before you own it. It's much harder to think clearly once you do.
5. How much of your portfolio would you commit, and why that amount?
Position sizing is a statement of conviction. Be honest with yourself.
Creates a clean one-page summary you can keep.
You have done the work.
This is your space to think it through.
Disclaimers
Important information about this research and how to use it

Not Personalized Investment Advice

The content of this report is impersonal analytical research. It is general in nature and does not take into account the specific investment objectives, financial situation, risk tolerance, or particular needs of any specific person who may read it. Nothing in this report constitutes an offer to buy or sell, or a solicitation of an offer to buy or sell, any security. Nothing in this report constitutes a recommendation that any particular security, portfolio of securities, transaction, or investment strategy is suitable for any specific person. The analyses and observations in this report are presented to help readers form their own judgments about whether the securities discussed are appropriate for their own circumstances. Readers should consult their own registered investment advisor, portfolio manager, or other qualified financial professional before making any investment decision based on information contained in this report.

Registration Status

GARPify is not registered as an investment dealer, portfolio manager, exempt market dealer, or investment advisor with any provincial or territorial securities regulator in Canada, with the United States Securities and Exchange Commission, or with any other securities regulator in any jurisdiction. GARPify operates as an impersonal financial publisher providing analytical research and commentary. GARPify does not provide personalized investment advice, manage client accounts, or hold client funds or securities.

Personal Holdings Disclosure

The author of GARPify research may from time to time hold positions in securities covered by GARPify reports. Specific holdings as of the publication date of this report: [To be completed by the author prior to publication: disclose by name any personal position held in LRCX, or state that no personal position is held.] The author may buy or sell securities covered by GARPify reports at any time, subject to the trading restrictions described in the GARPify Code of Conduct, which prohibits trading in covered securities for a specified blackout period before and after publication of new or updated research on those securities.

Source of Data and Information

Quantitative data in this report is sourced from YCharts and from primary regulatory filings retrieved from SEC EDGAR and SEDAR+. Specific data points are cited where they appear in the report. All data is current as of the publication date shown on the Cover tab. GARPify makes commercially reasonable efforts to ensure the accuracy of the data presented but does not warrant or guarantee the accuracy, completeness, or timeliness of any data, content, or opinions in this report. Readers should verify any specific data point against primary sources before relying on it for any investment decision.

Forward-Looking Statements

This report contains forward-looking statements about market conditions, business performance, valuation multiples, and potential investment outcomes. Forward-looking statements involve known and unknown risks, uncertainties, and other factors that may cause actual results to differ materially from any future results expressed or implied by such statements. The forecast scenarios in this report represent conditional outcomes under specific stated assumptions. They are not predictions. Actual outcomes may differ materially. Past performance is not indicative of future performance. Securities mentioned in this report can decline in value, and readers can lose money on investments in those securities.

Conflicts of Interest

GARPify is funded entirely by subscription revenue from its readers. GARPify accepts no advertising, sponsored content, paid placements, marketing support, or compensation of any kind from any company covered in its research, from any fund issuer, or from any intermediary that may distribute GARPify research. GARPify does not engage in investment banking, corporate finance advisory, brokerage, or any other business activity that would create a financial interest in the performance of securities covered in its research, other than the author's personal holdings disclosed above. GARPify research is selected, written, edited, and published solely by the GARPify author on an objective, documented methodology.