Conviction: High Status: Researching Listings: BESI.AS (Euronext Amsterdam, primary) | BESIY (US OTC ADR) | BESI (Madrid) Tickers used in this report: BESIY (the ADR — it's what our cache resolves; primary listing is BESI.AS)

2026-07-31 tape note: BESIY $229.53 (-10.43% vs $256.25), RSI 34.7 (cooled from 37.3). 7D -10.43% (deepened from -0.81%), 30D -27.88% (deepened from -22.75%), 3M -21.67% (deepened from -13.49%). From the 52-week high, the gap is -38.68% (widened from -30.54%). Per the tape: trend "weak-down", regime "pullback", golden cross intact. No new fundamentals re-researched. Per the full-scan summaries (2026-07-31 close). 2026-08-08 tape note: BESIY $256.00 (+11.53% vs $229.53), RSI 46.6 (warmed from 34.7). 7D +11.36% (turned positive from -10.43%), 30D -8.49% (moderated from -27.88%), 3M -17.13% (moderated from -21.67%). From the 52-week high, the gap is -33.38% (narrowed from -38.68%). Per the tape: trend "weak-down", regime "pullback", golden cross intact. No new fundamentals re-researched. Per the full-scan summaries (2026-08-07 close). 2026-08-14 tape note: BESI $270.96 (+5.84% vs $256.00), RSI 53.2 (warmed from 46.6). 7D +5.84% (moderated from +11.36%), 30D -4.21% (moderated from -8.49%), 3M -10.66% (moderated from -17.13%). From the 52-week high, the gap is -27.49% (narrowed from -33.38%). Per the tape: trend "weak-down", regime "pullback", golden cross intact. No new fundamentals re-researched. Per the full-scan summaries (2026-08-14 close).

---

Editorial Note

RSI cooling, thesis intact. BESIY is at $344, RSI 57.7 — cooled from the extended 65.8 RSI at the May 29 refresh. Price has continued grinding higher (+3.6% since May 29), now approximately -6.6% from the 52-week high (~$368.93). The RSI cooldown is constructive — the stock is digesting recent gains without breaking structure. The HBM + BSPDN cross-thesis convergence is unchanged; next catalyst watch is Q2-26 earnings (~late July 2026). Paradigm-validation pure-play thesis in good shape.

Paradigm-validation pure-play. Hybrid bonding is the cross-thesis tool of 2026 — the same equipment family sits inside both the HBM stacking ramp (memory side) and the BSPDN backside-power ramp (logic-foundry side), and BESI is the merchant Western leader on the D2W half of the toolkit. Our supply-chain-traces agent surfaced BESI months before NuttyCLD's "Last Micrometer" framed the cross-thesis convergence in plain English; that's three independent paradigms (autonomous trace, physics-down narrative, historical packaging-bottleneck conviction) all pointing at the same name. The wrinkle the market hasn't fully digested: BESI's Q4-25 orders were +43% sequential and +105% YoY, the SK hynix order for the AMAT+BESI Kinex inline system landed in March, and the next leg is foundry hybrid-bonding adoption on top of the HBM book — not instead of it.

---

The Story Right Now

For the first time in a while, our system is showing what cross-paradigm convergence is supposed to feel like. BESI was already in supply-chain-traces.json (autonomous-research agent discovery, April 11 HBM trace — flagged as the unanimous top-rank "85% market share in thermocompression bonding for HBM die stacking"). It was already implicit in the picks-and-shovels reading of the AI packaging shift. And now NuttyCLD's Part 4 — the "Last Micrometer" — explains why the convergence exists at the physics layer: BSPDN flips the chip and bonds it to a carrier wafer; HBM stacks die-on-die; both processes need atomically-precise hybrid bonding; the SAME family of tools serves both. SK hynix ordered the Applied Materials + BESI joint inline system (the "Kinex" platform) in March 2026 for its HBM line. That same family migrates next to BSPDN carrier bonding (W2W side, where EVG and Tokyo Electron lead) and to logic-foundry advanced packaging (D2W side, where BESI leads). The TAM doubles without BESI needing to win a single new socket — the equipment they already build gets pulled into a second adjacent ramp.

As of the June 16 refresh: BESIY is at $344, RSI 57.7, approximately +0.6% above SMA20 (~$342). The stock has continued its grind higher since May 29 ($331.95), up another +3.6%. RSI has cooled from 65.8 to 57.7 — a healthy digestion rather than a breakdown. The stock sits -6.6% from its 52-week high ($368.93). The entry zone at $260-280 cited in the original write-up is now well behind us. The bull case is playing out; the question remains whether to hold or take partial gains while the foundry BSPDN leg is still 12+ months out.

Sector context: RSI 57.7 is a neutral-bullish reading after the prior extended RSI run. The cooldown is constructive — digesting the +50%+ 3M gain without structural damage. A pullback from here to $320-330 (near SMA20) would still be normal and healthy. The pure-play nature and AMAT partnership remain structural advantages.

What makes this stock awkward is structural: it's a Dutch primary listing. BESI.AS on Amsterdam is the deep, liquid market; BESIY (the US ADR) is thinner — the leading-indicators short-interest data shows only 3,194 short shares, days-to-cover 1, short-volume ratio 47% (those numbers tell you the ADR float is small, not that the company is heavily shorted). The dividend ex-date was April 28, 2026 (€1.87 per ADR roughly), so any 30-day comparison sees that ex-div drop. P/E sits ~115-125x trailing TTM and ~54x forward — premium to AMAT/LRCX/ASML. The premium is the price of "pure-play hybrid bonding leader." That premium is the bear case in one number, and the bull case in two: orders.

Bottom line: the supply-chain-traces agent got there for the right physics reason (HBM packaging bottleneck), NuttyCLD got there for the right systems reason (the cross-thesis convergence with BSPDN), and Q4-25 results delivered the financial proof. The pure-play exposure to the bonding-equipment family is genuinely scarce — there's no real Western competitor at the same density of capability. The risk isn't whether the thesis is right; it's whether the timing of the foundry-side ramp lines up with current valuation, and whether you can stomach a Dutch listing with a thin US ADR. For paper personas this remains a HIGH-conviction watchlist name. At $344 RSI 57.7, hold existing positions; new entries closer to SMA20 (~$342) or on any pullback to $320-330 remain more defensible than chasing a new high.

TAPE-VS-THESIS [2026-07-11]: The "healthy digestion to $320-330" scenario named above has been exceeded to the downside. BESIY is now $293.34, RSI 41.4 — down -14.7% from this read, -20.43% over 30D, and -21.64% from 52wk high (was -6.6%). Price is now -10.5% below SMA20 ($327.64) and -7.9% below SMA50 ($318.45). This overshoots the "defensible new entry at $320-330" zone flagged above into the original $260-280 entry range's neighborhood. No thesis-breaking news identified since (pure price/RSI update); Q2-26 earnings (~late July 2026) remains the next real catalyst test of the HBM/BSPDN order book. Golden cross still intact.

TAPE-VS-THESIS [2026-07-17]: The slide has continued into the original entry zone. BESIY $257.23, RSI 32.5 (down from 41.4) — approaching oversold. -12.31% 7D, -29.91% 30D, -31.28% from 52wk high (widened from -21.64%). Now -16% below SMA20 and -18.8% below SMA50. Price is now inside the original $260-280 entry range (just below it). No new fundamentals re-researched; Q2-26 earnings (~late July 2026) remains the next real catalyst test — do not treat the RSI approach to oversold alone as a thesis confirmation ahead of that print.

TAPE-VS-THESIS [2026-07-25, Friday 2026-07-24 close]: The slide has stabilized inside the entry zone. BESIY $256.25, RSI 37.3 (up from 32.5, cooling out of approaching-oversold) — essentially flat vs the 07/17 checkpoint (-0.4%). 7D is now only -0.81% (vs -12.31%), and 30D has improved to -22.75% (from -29.91%). Still -10.4% below SMA20 (narrowed from -16%) and -18.2% below SMA50. 3M has deepened to -13.49% (from -5.32%, base-effect drag). -30.54% from 52wk high (roughly flat vs -31.28%). Golden cross vs SMA200 intact (+12.8%). No thesis-breaking news; Q2-26 earnings (~late July 2026) remains the gating catalyst — this cycle's tape reads as consolidation inside the original entry zone, not a fresh leg down.

DATA-FRESHNESS NOTE [2026-07-28]: No fresher BESIY tape is available — every watchlist that carries the ticker (ai-power.json, filing-watch.json, supply-chain-traces.json, wfe-test-metrology.json) still stamps most-recent data stamp: 2026-07-24, identical to the figures already captured in the 07/25 refresh above. This is a missing-data gap (BESIY's thin US-ADR data feed did not refresh over the 07/25-07/28 window), not a "no change" tape read — treat the 07/25 figures as the last known price/RSI/SMA state until a fresher pull lands. Q2-26 earnings (~late July 2026) remains the gating catalyst; no fundamentals re-researched.

---

Quick Snapshot

---

Action Matrix

---

Price Data

---

Company Overview

One-Liner

Dutch semiconductor assembly equipment maker — global leader in hybrid bonding (D2W) for advanced packaging used in HBM memory stacks and AI accelerator interposers; co-founded and run by Richard Blickman since 1995.

Business Model

Key Segments

Geographic Mix

---

Competitive Analysis

Industry Position

Competitor Map (the cross-thesis split)

Competitive Moat

Moat Assessment

Moat Width: Wide Moat Trend: Widening (BSPDN pulls a second demand curve onto the same toolkit; AMAT 9% stake locks in co-development)

Summary:

BESI sits at the intersection of two structural ramps with the same equipment family. The Kinex platform is in mass production at TSMC for AMD's 3D V-Cache and now ordered by SK hynix for HBM. ASMPT is the only credible Western-or-Asian D2W challenger and is one generation behind in HVM maturity. Switching costs in fab-qualified tools are years.

---

Management Assessment

Leadership

Founder Involvement

Capital Allocation

Red Flags

• Excessive exec compensation — none reported
• High turnover — opposite, founder still running it 30 years on
• Related party transactions — AMAT 9% stake is disclosed and strategic
• Aggressive accounting — Dutch listing, IFRS, conservative reputation

---

Financials

Single-source caveat (added 2026-08-03): no SEC-filed statements exist for this Dutch-listed name — figures below are company-reported and unreconciled.

Key Metrics (Q4-25 / FY-25)

Q1-26 Guidance

• Revenue: +5% to +15% vs Q4-25 (≈€175-191M)
• Gross margin: 63-65% (best-in-class for semicap)
• Opex: +10-15% (development spend continues)

Quality Checks

Revenue Quality

Leading Indicators

---

Valuation

Current Multiples

Valuation Context

• BESI's forward P/E ~54x ranks ~94th percentile in IT sector
• Premium reflects (a) hybrid-bonding pure-play scarcity, (b) AMAT/LRCX/ASML are too big for the bonding-cycle to move the dial, (c) ASMPT is the only direct comp and trades similarly
• The premium compresses when orders flatten — which is why Q4-25 +43% sequential orders is the load-bearing data point

Fair Value Estimate

My Fair Value (12-month): $345 — base case (P/E held flat, orders momentum continues into FY-26)

---

Bull Case

Why This Could Work

1. Twin-ramp tailwind on the same toolkit. HBM4/HBM5 stacking (D2W) and BSPDN carrier bonding (W2W) both pull on hybrid-bonding equipment. BESI leads D2W; the BSPDN ramp expands the category TAM and BESI gets cross-pollination on the carrier-bonding side too.

   • Evidence: SK hynix Kinex order March 2026; Intel Q1-26 advanced-packaging demand "billions per year vs hundreds of millions" expected
   • Implication: TAM doubles without needing share gains

2. AMAT 9% strategic stake + Kinex platform = locked-in customer. Applied Materials owns 9% of BESI and co-developed the Kinex inline system — the partnership is structural, not transactional.

   • Evidence: Kinex in HVM at TSMC for AMD V-Cache since 2022; SK hynix order 2026
   • Implication: AMAT's reach pulls BESI into every advanced-packaging conversation

3. No real Western pure-play competitor at the same density. ASMPT is the credible D2W challenger but is one generation behind in HVM maturity. EVG/Tokyo Electron play the W2W lane (different physics). K&S, Veeco, Aixtron are adjacent.

   • Evidence: ~85% market share thermocompression bonding for HBM (autonomous-research trace)
   • Implication: Pricing power per tool; gross margins 63-65% structurally

4. Pricing power on tools. Q4-25 GM 65%+; Q1-26 guidance reaffirms 63-65%. Best-in-class for semicap.

   • Evidence: Q1-26 guidance gross margin 63-65%
   • Implication: Operating leverage on the order ramp

5. Foundry BSPDN ramp 2027+ is the multi-year leg. Intel 14A maturity ahead of where 18A was; TSMC A16+SPR pushed to 2027 ramp; NVIDIA Feynman (2028) likely first AI accelerator on TSMC A16+SPR.

   • Evidence: NuttyCLD Part 4; Intel Q1-26 commentary; TSMC NA Tech Symposium April 2026
   • Implication: After HBM order leg, foundry hybrid-bonding orders become the next leg

Upside Scenario

---

Bear Case

What Could Go Wrong

1. Concentrated customer base = lumpy bookings. Top 5 customers = majority of revenue. One quarter of HBM digestion at SK hynix or one TSMC capex pause and orders flat-line.

   • How it plays out: Q1-26 orders disappoint vs Q4-25 €250M comp; stock sells 15-20% off RSI 77 levels
   • Probability: Medium — orders are lumpy by definition

2. TSMC A16 ramp slipped to 2027. Was guided to 2H-2026 volume production; April 2026 NA Tech Symposium confirmed slip to 2027. That's a real revenue-timing slip on the foundry-side leg.

   • How it plays out: Foundry hybrid-bonding orders that were 2H-26 expectations move out 6-12 months
   • Probability: Already happening — priced in partially

3. Dutch listing premium / thin US ADR. ADR illiquidity makes US institutional accumulation hard; primary listing on Amsterdam means EUR/USD risk on US ADR P&L.

   • How it plays out: Sharp moves on ADR with thin volume; harder for US buyers to size positions
   • Probability: High — structural

4. ASMPT closes the D2W gap. ASMPT is shipping 2nd-gen hybrid bonders to HBM customers Q3 2026. If qualification proceeds, BESI's D2W share erodes.

   • How it plays out: HBM5 socket split where today there's near-monopoly
   • Probability: Medium-low — first-mover advantage in HVM is durable, but worth watching

5. Valuation premium compression. Forward P/E 54x in IT-sector 94th percentile. Any orders disappointment and multiple compresses fast.

   • How it plays out: 30% multiple compression on order miss
   • Probability: Medium — the multiple is the bear case in one number

Thesis Killers

• ASMPT wins a major HBM5 socket from BESI at SK hynix or Samsung
• AMAT divests the 9% stake (would signal partnership cooling)
• Intel cancels PowerVia or TSMC abandons A16 BSPDN
• Two consecutive quarters of flat-to-down orders

Downside Scenario

---

Market-Moving News

Historical (What Has Moved BESI)

Recent News

News Patterns

Market cares about: (1) order momentum sequential, (2) major customer wins (SK hynix, Samsung, TSMC), (3) gross-margin sustain at 63-65%, (4) competitor disclosures from ASMPT. Overreactions cluster around quarterly orders prints — both up and down.

---

Related Markets & ETFs

Sector Context

The whole hybrid-bonding cohort is moving — ATEYY +27% 30D, KLIC +27% 30D, AMKR +56% 30D, ASX +43% 30D. BESIY +32% 30D fits the cohort move. The cross-thesis convergence is a cohort signal, not a single-name story.

---

Cross-References (Where This Appears in Our System)

---

Related Tickers to Monitor

---

Catalysts & Timing

Upcoming Catalysts

Key Dates

---

Entry Strategy

Position Sizing

My conviction: High (cross-paradigm validation is the strongest signal in our system right now) Target allocation: 4-5% (paper personas), 2-3% (real account, given thin US ADR + Dutch listing)

Entry Approach

Technical Levels

---

Sources

---

Research Log

---

The Gold

Key Discoveries

Open Questions

• Will SK hynix HBM5 socket split between BESI and ASMPT, or stay near-monopoly with BESI?
• What's the cadence of follow-on Kinex orders from SK hynix and prospectively Samsung/Micron?
• Does TSMC place a foundry hybrid-bonding tool order in 2H-26, signaling the BSPDN leg is starting?
• What's the right way to size BESIY in a real account given thin US ADR float? (consider BESI.AS direct via international broker for material size)

Calibration Notes

• Paradigm-validation: the supply-chain-traces agent (April 11) found BESI from a HBM-physics framing; NuttyCLD (April 25) found it from a BSPDN-physics framing; they converged on the same name through different paradigms. Log this as a calibration win in paradigm-scoreboard and convergence-db.
• Coverage gap closed: previously BESI (delisted) showed "no data" in supply-chain-traces summary. Update watchlist to BESIY so the cache populates going forward.
• Cross-thesis convergence: BESI is now the highest-conviction cross-thesis ticker in our system — appears in supply-chain-traces (HBM), ai-power-delivery (BSPDN), and optical-supply-chain adjacency (advanced packaging for CPO).