Back to News
Market Impact: 0.42

Applied Materials' China Problem Is Getting Worse

ACMR
AMAT
ASML
CTRYQ
ELST
INTC
MU
SKHYV
+5
Trade Policy & Supply ChainRegulation & LegislationSemiconductors & Supply ChainCredit & Bond MarketsCompany FundamentalsAnalyst Insights

China’s policy directing fabs adding new capacity to use at least 50% domestically produced semiconductor equipment is expected to structurally pressure foreign suppliers like Applied Materials over several years. While shares fell on reports that Shanghai Aishengna is producing ~5 immersion DUV systems in 2026, the article argues this is not a near-term earnings shock because Applied Materials does not manufacture lithography systems and the planned DUV quantities are small vs ASML’s 131 immersion DUV shipments in 2025. However, Applied Materials’ China exposure is large ($2.095B in fiscal Q1 2026, 29.9% of revenue), and indirect displacement risk rises as domestic equipment penetration at key growth fabs (e.g., CXMT with ~40%–50% domestic tool share) increases under the domestic-equipment requirement.

Analysis

The first-order read-through is misallocated: this is not a clean near-term earnings hit to AMAT, but a multi-year market-access issue for any tool vendor that still relies on China for secular growth. The immediate selloff in equipment feels like a forced de-risking of anything semiconductor adjacent; the better expression is that China is converting its largest fabs into a qualification lane for domestic tools, which compresses foreign vendors’ future share without necessarily showing up in next quarter numbers.

ASML has the clearest direct exposure because lithography is the bottleneck being localized, but the actual near-term unit count is too small to matter operationally. The more interesting second-order effect is that once a usable scanner exists, Chinese fabs can raise domestic content across deposition, etch, clean and CMP, which threatens the long-duration China annuity for AMAT and, selectively, ACMR. On the demand side, CXMT expansion is more bearish for MU and SKHYV over 6-18 months if it translates into bits, but the bigger immediate effect is import substitution, not memory oversupply.

Contrarian take: the market is right to price a structural de-globalization trend, but wrong to assume a 2026 P&L shock. The catalyst path is qualification data, not press releases: if SMIC/Hua Hong/CXMT actually run volume with domestic tools over the next 1-3 quarters, the China revenue mix of AMAT becomes progressively lower-quality. Falsifiers are simple: failed tool qualification, weak fab approvals, or any sign that China’s domestic content thresholds are being waived broadly for advanced nodes.