Chip Equipment Stocks Never Left the Top of Their Own Range

The companies that make chipmaking machines had a rough July. A report that China had started building its own lithography tools hit the group on July 27, in the middle of the worst month for chip stocks since 2008. It barely showed up in their valuations. As of October 5, 2026, Applied Materials, Lam Research and KLA trade at the 97th to 98th percentile of their own five-year EV/Sales range, at two to three times their own five-year medians. They started the year at the 100th.

If you only read one thing

The short version: Equipment makers are priced near the most expensive levels of their last five years and have been all year. The bull case is real: chipmakers keep raising spending plans, and Lam now expects the equipment market to reach the low $150 billion range in 2026. The risks are cyclical spending, China building its own tools, and multiples that leave no cushion.

The numbers

Ticker	Company	EV/Sales	5y median	EV/Sales %ile: Jan 5	July 27	Oct 5	P/E 5y %ile	Revenue growth (TTM)
LRCX	Lam Research	18.8x	6.2	100	96	98	98	26%
KLAC	KLA	20.4x	9.0	100	95	97	97	12%
AMAT	Applied Materials	13.9x	5.3	100	97	97	96	8%
ASML	ASML	n/a	n/a	snapshot only				21% (latest quarter)
8035.T	Tokyo Electron	10.3x	n/a	snapshot only				33% (latest quarter)

Percentiles rank each date's EV/Sales against that company's own weekly readings over the prior five years, computed from SEC filings and market prices. July 27 is the week of the July low. ASML files annual reports as a foreign issuer and Tokyo Electron lists in Tokyo, so we have current snapshots but no weekly history to rank. Trailing growth lags the latest quarter: Applied Materials' fiscal third quarter was up 25% and Lam's June quarter up 30%.

We use EV/Sales (enterprise value divided by trailing revenue) across the AI infrastructure theme because several members have no meaningful profit. For these three it barely matters: their P/E percentiles sit at the same 96th to 98th. Our guide to what a valuation percentile is and how we compute it covers the method.

The takeaway: On both sales and earnings, all three trade near the top of their own five-year ranges. July's drop in their prices left them there.

What the market is paying for

Equipment orders follow chipmakers' capital budgets, and those budgets went up through the summer.

• TSMC raised its 2026 capital budget to $60 billion to $64 billion in July and expects revenue growth slightly above 40% this year, per MarketBeat's summary of TSMC's second-quarter call.
• Lam raised its view of 2026 wafer fab equipment spending to the low $150 billion range, from $140 billion, and expects strong growth again in 2027, per the transcript of Lam's June-quarter call.
• ASML lifted its 2026 sales outlook to €43 billion to €45 billion and plans to add about 30% more EUV capacity in 2027, per ASML's second-quarter report.
• Applied Materials reported record revenue of $9.12 billion, up 25%, in its fiscal third quarter, and guided higher for the next.

The demand behind this is the AI build-out: leading-edge logic for accelerators, high-bandwidth memory, and advanced packaging that stacks the two together. Memory makers are a large part of the equipment order book; their own valuations are on the AI memory theme page.

The takeaway: Equipment makers are priced for a multi-year upcycle, and every major customer has guided to one. The question is how much of 2027 is already in the price.

The China question

On July 27, reports said a state-backed Shanghai company had started producing immersion DUV lithography machines, the workhorse tool for older chip generations. Output is small, about five machines this year and about 20 next year, and some parts are still imported, according to TrendForce's summary of the reports. ASML slid on the news, per Bloomberg's report on the ASML selloff, and US equipment makers followed.

China matters because it bought heavily in 2024 and 2025. Lam's China share of revenue fell to 26% in the June quarter from 34% the quarter before, per its call. For now, the order books have grown faster than China has shrunk.

The takeaway: Chinese lithography is early and small, but it targets the mature-chip spending that carried much of the equipment makers' China sales. It is a slow threat, not a sudden one.

The bear case

• Spending is cyclical. Equipment revenue rises and falls with chipmakers' budgets. A cut from TSMC, a memory maker or a cloud company flows straight to orders.
• Multiples have no cushion. At two to three times their own five-year median EV/Sales, these stocks price in years of growth. A flat 2027 would leave them with a long way to fall within their own ranges.
• Export rules cut both ways. Tighter US limits on sales to China hit revenue; looser ones help Chinese chipmakers build capacity that competes with the equipment makers' other customers.
• Trailing growth is uneven. Applied Materials' trailing revenue grew 8%, the slowest of the three, even though its latest quarter was up 25%.

What to watch

• 2027 equipment forecasts. Lam's "extraordinary setup" needs to show up as guided numbers in the January reports.
• TSMC's 2027 capital budget, usually given in January.
• China's share of revenue at each company, and any volume news on Chinese lithography tools.
• The percentiles themselves. If they hold near the top through a quarter of flat orders, the market is looking past the cycle. If they break, the cycle is back.

Our ranking of AI infrastructure stocks by valuation percentile puts the equipment makers next to the chip designers, networking suppliers and neoclouds they serve, and Nvidia and AMD's opposite readings show how differently the market prices the buyers of TSMC's output. Each equipment maker has its own band chart: Applied Materials' EV/Sales history, Lam Research's EV/Sales history and KLA's EV/Sales history. The power side of the build-out is on the AI power theme page.

---

Data: SEC EDGAR XBRL filings and market prices, computed by StockResearch as of October 5, 2026. Earlier-date percentiles were computed from the same weekly series using only the five years available at each date. Latest-quarter growth figures come from company releases and calls. Jake is the founder of StockResearch.app, where he writes data-first research on valuation dislocations. Nothing here is investment advice or a recommendation to buy or sell any security. Do your own research.