Bernstein Raises Global WFE Spending Forecasts for 2026-2028
nashnova research
Bernstein raised its 2026–2028 global wafer fab equipment forecasts to $156B, $217B, and $279B respectively, arguing that the capex cuts the market feared have not materialized — memory expansion and packaging demand are pushing the equipment cycle toward its strongest three-year window in a decade.
How much was raised, and where is the money coming from?
Bernstein lifted its 2026 global WFE — wafer fabrication equipment, the full suite of machines used to make chips — forecast from $148B to $156B, implying 28.0% year-on-year growth.
2027 moves from $204B to $217B (+39.3%); 2028 from $259B to $279B (+28.4%).
This means → a cumulative ~$40B upgrade across three years, driven by higher spending from Chinese and overseas logic/foundry fabs, DRAM makers, and upward revisions to wafer-level packaging (WLP) equipment demand.
Why is memory the main engine of this cycle?
DRAM equipment spending is forecast to grow 57% and 39% year-on-year in 2027 and 2028; NAND at 58% and 42%; logic/foundry at 27% and 19%.
In plain terms = memory fabs are expanding at roughly twice the pace of logic fabs — the bulk of equipment orders sits on the memory side.
WLP packaging equipment was also sharply revised up: +$2B in 2026, +$4.4B in 2027, +$7.9B in 2028. This reflects advanced packaging shifting from a supporting role to an independent growth driver in equipment spending.
Where is China spending?
China WFE demand holds at $57B for 2026; 2027 is raised $2.9B to $76B; 2028 raised $6.7B to $108B.
The core driver: leading Chinese memory makers have firmed up expansion plans — China memory WFE is projected to grow 84%, 78%, and 68% year-on-year from 2026 to 2028.
On the advanced-logic side, breakthroughs such as Huawei's logic folding and AI networking sustain demand. In mature nodes, AI infrastructure buildout and auto-chip localization may tighten capacity, pushing Chinese fabs to accelerate expansion in 2027–2028.
How does Bernstein rank the overseas equipment makers?
Applied Materials, KLA, and Lam Research all carry outperform ratings. Ranking: Applied Materials > KLA > Lam Research.
This means → Bernstein favors Applied Materials for its broadest product portfolio and lowest relative valuation; KLA for a favorable growth base and strong 2027 earnings leverage; Lam for execution and margins.
In Japan, the preferred order is Kokusai Electric > Tokyo Electron > Screen. Kokusai leads in batch ALD — atomic layer deposition, a technique that deposits thin films one atomic layer at a time — and benefits deeply from NAND and GAA node transitions. In Europe, ASML is the top pick at a €2,500 target; EUV shipments are expected to rise from 48 units in 2025 to 118 in 2028.
Is now the time to buy Chinese equipment stocks?
Bernstein maintains outperform on NAURA, AMEC, and PIOTECH, ranked: NAURA > AMEC > PIOTECH.
Parts-shortage delivery delays are expected to ease over the next one to two quarters; the recent pullback is seen as a positioning window, with an earnings recovery projected for 2027.
In plain terms = NAURA has the most complete product matrix — etch, deposition, thermal, cleaning — and the widest customer base, giving it the highest certainty of benefit. AMEC leads in dry etch and is expanding into deposition; PIOTECH focuses on thin-film deposition and advanced packaging equipment. Bernstein sees no slowdown in NAND expansion; after YMTC's IPO restarts, AMEC and PIOTECH could see an earnings rebound as early as Q1 2027 — a key checkpoint for validating the upgrade thesis.
市场有风险,内容仅供研究参考,不构成投资建议。
