Exclusive-Chinese chipmaker YMTC plans new factories amid heightened US-Sino trade tensions, sources say

YMTC plans two additional factories plus one nearing completion, which would more than double capacity to over 350,000 wafers per month when fully operational. The Wuhan third fab is due to start late this year and reach 50,000 wafers per month by 2027, with more than 50% of equipment sourced domestically amid U.S. export restrictions. The expansion underscores China’s push to reduce reliance on foreign semiconductor technology and could support YMTC’s NAND market share, which UBS estimates at 11.8% last year and above 14% by early 2027.

Analysis

The key second-order effect is not just more Chinese NAND supply, but a faster substitution cycle away from foreign memory and equipment ecosystems. If YMTC can keep scaling with domestic tools, the real pressure shifts to global vendors selling high-end process equipment, process control, and memory IP into China rather than to memory pricing alone; that means export-control risk is becoming a recurring earnings overhang, not a one-time headline.

For memory competitors, the incremental capacity is most damaging to the mid-cycle recovery thesis. A larger YMTC footprint can cap NAND upside just as inventories normalize, because China’s demand base is large enough to absorb local output before imports rebound, which compresses the effective addressable market for names like SNDK even if spot pricing firms. The bigger medium-term issue is that DRAM optionality creates a path for China to attack two memory markets at once, increasing the probability of policy escalation from Washington if domestic substitution accelerates faster than expected.

On UBS, the direct read-through is more about flows and sentiment than fundamentals: an improving China memory stack tends to reduce the market’s perceived moat around Western coverage/financing franchises, while also increasing volatility in Asia tech risk appetite. The contrarian point is that this may be underpriced as a structural rather than cyclical change; if domestic tool sourcing keeps rising, each additional restriction may speed localization rather than slow it, leaving foreign suppliers with lower China revenue but not necessarily lower Chinese capex. The main timing catalyst is policy over the next 1-3 months, but the strategic earnings impact plays out over 12-24 months.

Key Decisions for Investors

• Short SNDK on a 3-6 month horizon into strength; thesis is that incremental Chinese supply plus localization weakens the NAND pricing recovery and caps multiple expansion. Use a defined-risk structure if available, since any near-term AI/storage demand spike can squeeze the trade.
• Pair trade: long non-China exposed memory/enabling names vs short China-levered suppliers/beneficiaries of policy tightening. Best expressed as long quality semiconductor equipment beneficiaries outside China and short a basket of China memory proxies if liquid.
• Buy 6-12 month out-of-the-money call spreads on U.S. chip-equipment names as an export-control hedge; further restrictions raise order visibility from ex-China fabs and can offset China capex leakage.
• Fade any rally in UBS tied to Asia tech optimism; use it tactically as a volatility expression rather than a directionally strong long, since the market may be underestimating how persistent export-control headlines keep China-related risk premiums elevated.