EWN joins the run: equipment wakes up one week after the foundry breakout

The structural change since the May 20 entry: EWN (Netherlands/ASML) has moved from -0.53% to +6.5% over the rolling 30-day window. One week ago the thesis was "equipment lags foundry by 12-18 months." That lag is now compressing at the price level. The rest of the chain has also accelerated: EWY is +29.81% 1m (was +15.94%), EWT is +15.95% (was +7.47%). The whole AI value chain is moving in parallel now, not sequentially.

What I'm watching

• EWN (Netherlands): $67.33, 1m +6.5%, 52w-pos 100%. ASML is at its 52-week

high. On May 20 it was flat while foundry ran. The breakout in equipment is the single most meaningful update this week.

• EWY (Korea): $200.65, 1m +29.81%. The Samsung HBM3e + KUSPI bilateral +

governance re-rating thesis is still the structure, but a 30% monthly move raises the question of how much forward earnings are priced in.

• EWT (Taiwan): $102.14, 1m +15.95%, 52w-pos 100%. TSMC at 52-week high;

no tactical entry signal, but thesis intact as structural hold.

• Copper: 6.402 USD/lb, 30d +5.43%. Physical AI infrastructure demand (power,

cooling, interconnect) is still expanding, not decelerating.

What changed / what matters

EWN breaking out: reading the equipment signal

The May 20 entry explained EWN's flatness as correct market behavior: equipment orders lag foundry utilization by 12-18 months. Current-capacity yield is being harvested; new-node capacity investments come later. That logic is still valid for the multi-year horizon. But at the monthly price level, EWN at +6.5% suggests the market is now pricing something forward. The most plausible mechanism: ASML's forward order book visibility on high-NA EUV is starting to harden into analyst earnings revisions. High-NA EUV tools have lead times of 18-24 months; if TSMC and Samsung are signaling commitment to the next node (N2 derivatives, 1.4nm roadmap), ASML's 2027-2028 revenue line becomes more legible, and the stock re-rates ahead of the delivery cycle. This is the normal equipment-cycle dynamic, arriving roughly on schedule relative to when EWT started running in early 2026.

EWN at 52w-pos 100% removes the "undervalued vs chain" argument. It is no longer a patient hold priced for discount. The holding argument shifts to: the high-NA EUV moat is durable (zero competitors, 8-10 year replacement cycle), and the next major capacity commitment would extend the order backlog materially.

EWY: thesis intact, momentum now dominant

At +29.81% 1m, EWY's price action is increasingly momentum-driven on top of a still-valid thesis. The three legs identified on May 20 (governance re-rating, KUSPI bilateral anchor, HBM3e cycle recovery) have not reversed. But a 30% monthly move in a single-country EM ETF is the kind of number that attracts systematic rotation into it from quant strategies, creating reflexivity. The risk is not that the thesis is wrong, but that a position entered here has compressed the margin between thesis value and current price. DCA pacing is the mitigation, not exit.

India: compound-semiconductor buildout does not move the ETF

India's ISM Phase 1 final tranche (IPTM 2026-05-05-india-ism-crystal-matrix-suchi-semicon-cabinet-approval, severity 3) approved CML's GaN/Micro-LED fab and Suchi Semicon's OSAT facility in Gujarat -- INR 3,936 crore, closing Phase 1 before ISM 2.0 takes over. The GaN angle is analytically notable: India is building compound-semiconductor capability in a space where China holds over 70% of global display-panel production. That addresses a real supply-chain derisking vector for OEMs. But INDA remains 70%+ weighted to software services (TCS, Infosys, Wipro, HCL), so ISM investments do not translate to INDA returns on any near-term horizon. INDA's -2.08% 1m (improved from -6.45% on May 20) is still a software-margin-compression story, not a semiconductor-buildout story. The structural thesis strengthens; the ETF vehicle still has the wrong composition to capture it.

Candidate picks within this theme

• EWY (Korea) -- thesis intact across three legs; momentum extended but not

broken; DCA pacing more important than ever at +30% 1m.

• EWT (Taiwan) -- TSMC node leadership at 52w high; no new entry, structural hold.
• EWN (Netherlands) -- equipment cycle joining the foundry run; ASML high-NA EUV

moat is durable; no longer discounted, but the forward order visibility argues for hold.

• EWJ (Japan) -- equipment diversification (TEL, Disco, Advantest) plus BoJ

normalization; lowest-beta way to stay in the chain.

• INDA -- omit for tech-ai theme; composition mismatch persists despite ISM buildout.

What I'd revise if I saw

• ASML booking guidance cut at the next earnings call: signals hyperscaler capex

pulling back, reprices EWN and compresses the EWT/EWY equipment-dependent layer.

• Samsung HBM3e qualification confirmed at disappointing volume (below 2mn units

per quarter): reduces EWY from multi-thesis to single-theme, likely drops 2 slots.

Cross-references

• Previous entry on this theme: docs/thinking/2026-05-20-tech-ai-korea-breakout-foundry-gap.md
• India ISM final Phase 1: docs/iptm/actions/2026-05-05-india-ism-crystal-matrix-suchi-semicon-cabinet-approval.md
• Relevant macro routes: /api/country/KR, /api/country/TW, /api/country/JP, /api/country/IN
• Current picks: docs/thinking/2026-05-23-saturday-synthesis.md