The tools that build every advanced chip have never sold better. The stock has rarely been priced to forgive so little.
Lam prints record revenue and a record 50% gross margin as AI rebuilds the semiconductor industry layer by layer — yet after a +255% year to an all-time high, LRCX just shed ~20% in days on one question: is this a durable re-rating of Lam’s earnings power, or a cyclical peak a ~66× multiple has mispriced? Five analyst lenses, three scenarios, four horizons.
Gray line = LRCX’s actual split-adjusted price ($91 low mid-’25 → $438.50 all-time high Jun 30 ’26 → $351.41 now after a sharp AI-trade reset); colored paths = synthesized scenario midpoints forward, probability-weighted (base 45% · bear 30% · bull 25% — bear-tilted because the multiple, not the business, is elevated). Log-linear, mid-year marks. Wall Street 12-month consensus ≈ $349 (range $220–$500) now sits right at the tape. 10-for-1 split effective Oct 3, 2024; all prices post-split.
Re-weight the scenarios
Those probabilities are a judgment call — so make them yours. Drag to set how likely the bear and bull cases are (base takes the remainder); the blended target below, the dotted line on the chart, and the prob-weighted row of the scenario cards all update live. The default leans bearish (30/45/25) because a ~66× multiple, not the business, carries the risk.
Five analyst lenses, five answers
The same record fundamentals justify wildly different conclusions depending on which framework you trust — and on a stock up 255% in a year, framework is the whole argument. Each lens below is a synthesized expert perspective with its own 12-month target.
The Layer Multiplier
Every AI chip is built by depositing and etching more, thinner layers — so Lam’s tools matter more each node, not just more often. Management’s served share of WFE has climbed from the low-30s to the mid-30s toward high-30s; advanced-packaging revenue is guided to grow 50%+ in 2026. As chips go 3D, intensity outruns unit volume. Record EPS +41% deserves a premium.
The Cash Compounder
A ~50% gross margin, ~66% return on equity, and 139% of free cash flow returned in the quarter — buybacks (with $4.3B left) plus a dividend. This is a rare quality-and-cash business. But at ~66× earnings the downside floor is thin: quality doesn’t rescue you from a multiple that assumes nothing ever cools.
The Peak-Cycle Short
WFE is cyclical and ~66–80× earnings is more than double Lam’s 5-year average of ~25×. System-unit growth is decelerating sharply beneath the record ASP-driven headline; the CEO sold ~30,000 shares near the high. If OpenAI-style efficiency and hyperscaler overbuild trim tool orders, earnings and the multiple compress together — the double de-rate.
The Etch Monopolist
Lam is #1 in etch — the largest WFE segment — and clear #2 in deposition, the two steps that scale with 3D structure. At the leading edge these tools are effectively single-sourced, and Equipment Intelligence / Dextro automation deepens the switching cost inside customer fabs. The moat is real; the question is only what you pay for it.
The Momentum Unwind
+255% in a year, an RSI stretched to extremes, and a clean break below the June breakout shelf as the whole SOX fell 6% in a session. Beta ~2.3 means LRCX amplifies every AI-sentiment swing. With the mean Street target now below the price, the reflexive momentum that drove the run can reverse just as fast.
Wall Street 12-month price targets
What the sell-side expects over the next year. Bars are sorted low to high and colored by rating; the dashed line is today’s $351.41. The tell: after the run, consensus (≈$349) has been dragged up to meet the price — the euphoria is priced out, and the debate is now two-sided.
Sell-side 12-month targets — a selection of the ~41 firms covering Lam; the full consensus is ≈ $349, roughly flat to today, with a still-bullish skew (about 29 buy / 6 hold / 1 sell). Recent hikes ran hard (Cantor $500, BofA $480, Susquehanna $475), but the low end ($220) implies a real de-rate. The dashed line at $351.41 shows how far the stock has run: it now sits above the average target, so upside from here needs the bull case, not just consensus. Firms, ratings, and targets illustrative.
Where the layers lead
Synthesized scenario midpoints (mid-year). Returns shown vs. today’s $351.41. These are illustrative frameworks, not predictions — and for a cyclical, the range is genuinely wide: the bull compounds a durable super-cycle, the bear rides a spending pause and a multiple that halves.
1 Year
Mid-20272 Years
Mid-20283 Years
Mid-20295 Years
Mid-2031▸ Bull case — show the assumptions & math
▸ Base case — show the assumptions & math
▸ Bear case — show the assumptions & math
Revenue, capex, free cash flow & debt ($B)
Where the money actually goes — on a June fiscal year. The whole bull/bear split lives in one shape: revenue dipped in the FY24 down-cycle, then AI drove it vertical, while capex stayed tiny and debt fell. Is that vertical line structural, or the top of a cycle?
Lam’s asset-light engine in one view: revenue fell to ~$14.9B in the FY24 down-cycle, then AI drove it toward ~$23B in FY26E — and free cash flow (olive) rose with it toward ~$5.8B. Capex (clay) is strikingly small — roughly 5% of revenue — so almost all the cash is free, funding buybacks and a falling debt load (slate), which the $750M note retirement pushed lower still. The bull sees a cash machine; the bear sees a revenue bar that has only ever gone vertical at the top of a cycle. Figures illustrative; fiscal years end late June, debt is gross, FCF is annual.
The EPS ladder underneath the targets ($)
The price targets aren’t pulled from the air — each is an EPS estimate times an exit multiple. Here’s the earnings ladder the scenarios are built on, and note the FY24 dip: this is a cyclical business, which is exactly why the exit multiple matters as much as the earnings.
Non-GAAP EPS, split-adjusted. Gray = reported (note the FY24 down-cycle dip from $3.40 to $3.05 — the reminder that WFE is cyclical); olive = estimates assuming AI-driven growth compounding through the back half of the decade. The base case’s ~$10.3 of FY2031 EPS at a ~54× exit multiple ≈ the $560 base-case 5-year target; the bear applies a normalized ~31× to a lower through-cycle number. This ladder is what literally sits under the prices. Estimates illustrative, not guidance.
The business is firing on every line
Q3 FY26, year-over-year (advanced packaging is management’s 2026 growth guide). Read these against a stock that just fell 20% — if the fundamentals are this strong, the sell-off is a statement about the multiple, not the machines.
Every line is green — revenue +24%, operating income +35%, EPS +41% — with the AI-specific frontier line, advanced packaging, guided to grow 50%+ in 2026 (clay), alongside record DRAM/HBM mix. The catch the bear presses: system-unit shipments are decelerating hard beneath these ASP-and-mix-driven percentages, so the headline growth may not repeat. That disconnect — strong prints, a falling stock — is the whole valuation argument in one chart. Frontier figure is a full-year management guide, not a Q/Q print.
Bull vs. Bear
The entire valuation argument compresses into one disagreement: is Lam’s AI surge a structural step-up in through-cycle earnings power — or a cyclical peak that a ~66× multiple has already over-paid for?
▲ THE BULL CASE
- Records across the board. Q3 FY26 revenue $5.84B (+24%), non-GAAP EPS $1.47 (+41%), ~50% gross margin — and the June guide accelerates to $6.6B / $1.65.
- Intensity beats volume. As chips go 3D (gate-all-around, HBM, advanced packaging), each node needs more etch and deposition — Lam’s served WFE share has risen from the low-30s to the mid-30s, targeting the high-30s.
- The end market is growing. Management raised 2026 WFE to $140B with “a bias to the upside,” sees visibility into 2027, and flags ~$40B of NAND-conversion spend pulling forward.
- An etch near-monopoly. #1 in etch (the biggest WFE segment) and #2 in deposition; at the leading edge the tools are effectively single-sourced, and automation deepens the lock-in.
- A cash machine. ~66% return on equity, 139% of free cash flow returned in the quarter, $4.3B of buyback authorization left — retiring shares even as it retires debt.
- Recurring ballast. The Customer Support (CSBG) business crossed $2.1B in a quarter (+25%), a growing installed-base annuity that softens the cycle.
▼ THE BEAR CASE
- Priced for perfection. ~66–80× earnings is more than double Lam’s 5-year average of ~25× — the multiple, not the business, is what a buyer is really underwriting.
- WFE is cyclical. Wafer-equipment spending has always peaked and troughed; memory (NAND/DRAM) capex is lumpy and front-loaded, so a pause hits orders hard.
- The AI-capex-peak fear. OpenAI-style efficiency gains and hyperscalers (Meta) reselling excess capacity raise the question of whether the build-out is closer to a peak than a floor.
- Units are decelerating. Beneath the record ASP-and-mix-driven revenue, analysts see system-shipment growth slowing sharply from the prior year’s pace.
- China = 34% of revenue. A large slice sits under active U.S. export-control risk that can be tightened with little warning.
- The double de-rate. If demand normalizes, earnings and the multiple compress together — and the CEO selling ~30,000 shares near the high didn’t help sentiment.
Risk map — likelihood × impact
Where each risk sits, not just how big it is, over a 3–5 year horizon. The hot upper-right — likely and high-impact — is the one that matters, and for LRCX it isn’t a competitor: it’s the multiple itself.
- Insider selling / sentiment
- Memory (NAND/DRAM) lumpiness
- Valuation de-rating
- China export controls
- Customer concentration
- WFE cycle rollover
- AI capex peak
- Geopolitical / supply shock
- Technology disruption
Valuation de-rating
The core risk isn’t a rival — it’s the ~66× multiple. A normalization back toward the ~25–30× historical average roughly halves the stock even if earnings simply hold flat.
WFE cycle rollover
Wafer-equipment spending is cyclical; if memory and foundry capex pause rather than compound, orders soften and the premium multiple compresses at the same time.
AI capex peak
OpenAI-style efficiency gains and hyperscaler overbuild (Meta reselling capacity) could cut the tool orders that the whole AI-WFE thesis rests on.
Geopolitical / supply shock
A Taiwan-Strait or Korea disruption — where Lam’s leading-edge customers concentrate — would freeze demand and reprice the platform overnight. Low odds, severe consequence.
Technology disruption
A patterning or architecture shift that reduces etch/deposition intensity would erode the very trend the bull case is built on. Unlikely near-term, but existential if it lands.
China export controls
China is ~34% of revenue; tighter U.S. rules on advanced-node tools can remove a chunk of sales with little warning.
Memory (NAND/DRAM) lumpiness
Memory spending is Lam’s most volatile end market and much of the NAND-conversion spend is front-loaded — a natural air-pocket after a pull-forward.
Customer concentration
A handful of leading-edge foundry and memory customers drive most systems revenue; one deferring capex swings a quarter.
Insider selling / sentiment
The CEO sold ~30,000 shares near the record; a modest fundamental signal, but one that feeds a jittery, high-beta tape.
The jargon, decoded
Hover the dotted terms in the metrics, or scan the desk’s working definitions here.
- WFE
- Wafer fabrication equipment — the machines chipmakers buy to build chips. Lam’s total addressable market; guided to ~$140B in 2026.
- Etch & deposition
- The two process steps Lam leads: deposition lays down thin material layers, etch selectively carves them away. Both scale with 3D chip complexity.
- Served share of WFE
- The slice of total equipment spend Lam’s products can address. Rising from the low-30s toward the high-30s% as chips get more complex.
- HBM
- High-bandwidth memory — stacked DRAM that feeds AI accelerators. Its build-out is deposition- and etch-intensive, a key Lam tailwind.
- Advanced packaging
- Stitching multiple chips into one package (chiplets, 3D stacks). A fast-growing, tool-hungry step Lam guides to grow 50%+ in 2026.
- Gross margin
- Revenue left after the cost of building the tools, as a percent of revenue. Lam’s ~50% signals strong pricing power.
- Free cash flow
- Cash left after running and investing in the business — the fuel for buybacks and the dividend. ~$5.7B trailing.
- Exit multiple
- The P/E assumed at the end of the forecast. Multiply it by projected EPS to get a target price — the swing factor for a stock at ~66×.
- De-rating
- The multiple falling even if earnings hold — the bear’s core mechanism here: ~66× drifting back toward a ~25–30× norm.
- Prob-weighted
- Each scenario’s price × its probability, summed into one expected value across bear, base and bull.