India's Capex Boom Is Not Driving Semiconductor Equipment Stocks

India's largest bank, State Bank of India, says a new-age capex cycle is under way. Data centers, green hydrogen, electric vehicles, and semiconductor manufacturing are the drivers. SBI reports credit demand "very strong" and bank lending up 19% year-over-year through July. The headline is bullish. The question for an investor who holds or watches U.S. semiconductor equipment stocks is whether this Indian investment wave flows into the companies they own.

The answer, by the numbers, is no. Not yet, and possibly not for a decade. The gap between the announcement scale and the actual equipment content of India's capex program is the material finding.

What India Is Building

The India Semiconductor Mission has approved roughly a dozen manufacturing and packaging facilities across six states. The headline project is Tata Electronics' $11 billion wafer fab in Dholera, Gujarat, partnered with Taiwan's Powerchip Semiconductor Manufacturing Corp and supported by an ASML agreement signed in May 2026. But the node specification matters. Tata's fab will produce 28nm and above — mature nodes used in automotive chips, power management, display drivers, and IoT devices. India will not have a leading-edge sub-7nm fab before 2030 at the earliest.

The rest of the pipeline is even further from the leading edge. Micron's OSAT facility in Gujarat is already producing DRAM and NAND assembly and test. CG Power, Tata's TSAT, Kaynes, HCL-Foxconn — these are assembly, test, and packaging operations. OSAT facilities require equipment, but the capital intensity is a fraction of that of a full wafer fab. A 300mm fab can require $10 billion to $20 billion in equipment alone. An OSAT line typically needs a few hundred million to a couple of billion.

India is simultaneously pushing a data center pipeline exceeding $100 billion, with Adani planning $100 billion in AI-ready data centers by 2035 and Reliance pledging up to $110 billion over seven years. A new law exempts foreign cloud providers from tax until 2047 if they use Indian-owned data centers.

The Equipment Disconnect

The U.S. investor should ask which companies sell into this pipeline. Then ask whether those companies' stock prices are already reflecting something entirely different.

ASML signed the Tata deal, and that is the most direct India-equipment link in the industry. But ASML's Q2 2026 financials — €9.3 billion in net sales, raised full-year guidance of €43 billion to €45 billion, gross margin expanding to 54%–56% — are driven by EUV and high-NA EUV shipments to Taiwan and South Korea. Taiwan accounted for 30% of ASML's Q2 sales share, up from 23% a year ago. South Korea held 43%. China was under 20%. India does not appear in the breakdown.

The Tata fab will use ASML's DUV tools, not EUV. DUV is ASML's lower-margin, volume business. Even at full ramp — 50,000 wafers per month — the equipment content of one mature-node fab, even at $11 billion project cost with roughly 50% government subsidy, represents a rounding error against ASML's €43 billion-plus annual revenue run rate.

Applied Materials, Lam Research, and KLA face the same structural gap. These companies are seeing revenue acceleration — Applied Materials grew Q2 revenue 13% year-over-year to $7.9 billion, Lam Research posted 26% annual revenue growth — but the driver is leading-edge capacity expansion in the U.S., Taiwan, and South Korea, not India. Lam Research announced a $1.15 billion investment in Karnataka, but that is a manufacturing and service center for Lam's own operations, not a sale of tools to Indian fabs.

The Maturity Problem

There is a second, more consequential layer. The segment India is building into — mature nodes, 28nm and above — is exactly where the industry faces the most pricing pressure.

After the 2022–2023 semiconductor downturn, leading-edge suppliers (TSMC, Samsung) exercised supply discipline. They constrained expansion, prioritized node migration, and let pricing recover before volume. That supply discipline is what drove the memory and foundry recovery of 2024 and 2025.

But mature nodes have a different supply dynamic. China's domestic foundries — SMIC, Hua Hong, and newer entrants — have been expanding aggressively at 28nm and above, insulated from Western export controls at these nodes. The result is a growing pool of mature-node capacity that competes on price, not technology.

India's new fabs, subsidized at up to 50% of eligible project cost by the government, enter this competitive environment with a structural advantage that comes from a structural vulnerability. The subsidy lowers the effective cost of capacity. But the output competes in a global market where Chinese foundries are also subsidizing capacity. The economics work for the Indian government's strategic objective — supply chain diversification and import substitution. They may not work as well for the companies that bear the operating cost once the subsidy window closes.

Where the Real Constraint Is

The capex cycle that is actually driving semiconductor equipment valuations today sits at the leading edge. ASML is adding 30% more EUV and DUV capacity in 2027 and studying another 30% increase for 2028. TSMC reported a 68% jump in June sales and announced two additional advanced packaging plants in Taiwan. The foundry expansion that matters for ASML, Applied Materials, and Lam Research is happening in Arizona, Phoenix, Taiwan, and South Korea.

India's cycle matters for Indian industrial policy, for the local supply chain, and eventually for equipment vendors once the fabs come online between 2027 and 2030. But it does not explain the current price of U.S. semiconductor equipment stocks. Those prices are set by AI-capacity demand, leading-edge node migration, and the willingness of hyperscalers and governments in the U.S., Taiwan, and South Korea to spend tens of billions on sub-7nm production.

The implication is fairly straightforward. The India capex story is real. It is not the driver of the equipment stocks most U.S. investors own, and it will not be for years. If you are evaluating ASML, Applied Materials, or Lam Research, the India narrative is context — not causation.