Deck
ASMPT is a Hong Kong-listed maker of back-end semiconductor equipment — bonders that assemble the dies inside AI memory and AI-logic packages, plus surface-mount placement systems now under formal strategic review.
Stock is paying ~80× trough earnings for a business earning a third of BESI's margins.
- The multiple. HK$176 prints at ~80× FY2025 P/E and 5.2× EV/sales after a +229% trailing year. That puts ASMPT inside a notch of pure-play BESI (80× P/E, 35× EV/sales) and Hanmi (50× EV/sales).
- The economics underneath. 5–7% operating margin, 3.6% ROA, 37.8% gross margin. BESI prints 29% operating margin, 29% ROIC, 63% gross margin on a purer advanced-packaging mix.
- What has to be true. Consensus FY2026 EPS of HK$4.01 requires +258% YoY off a trough and sustained gross margin above 40% as TCB mix dominates. One quarter below 38% lands the multiple closer to broad-line peers (KLIC 6.8× EV/sales) than to BESI.
Sole-supplier qualification at TSMC CoWoS is real. The hybrid-bonding gap to BESI is the part that ages it.
- Process-of-record stickiness. Sole supplier of chip-to-substrate TCB at TSMC's CoWoS line, with HBM4 12-high process-of-record at multiple HBM players confirmed in February 2026 — breaking the prior 'Hanmi anchors SK hynix' narrative. Each qualified package locks 5–7 years of cumulative orders.
- Installed-base annuity. 500+ TCB tools shipped cumulatively — the largest base globally. Each tool throws off 7–12 years of spares and upgrades. SEMI gross margin held above 43% even at the FY2024 revenue trough.
- The weak spot. HBM5 (~2028-29) shifts the industry from TCB to hybrid bonding. BESI has shipped 150+ hybrid bonders to 18 customers and installed the first integrated production lines. ASMPT's gen-2 LITHOBOLT platform has no comparable customer count disclosed — a 24-month gap that defines the long-run rating.
Reported profit looks like a recovery. The cash flow says it hasn't arrived yet.
Headline EPS more than tripled to HK$2.61 on continuing ops — but HK$1,114M of HK$1,216M in pre-tax profit was the one-off AAMI joint-venture disposal gain. Strip restructuring, impairments and the JV and underlying PBT was HK$586M versus HK$559M (+4.8%), not the +167% headline. Management's own adjusted continuing net profit (HK$467M) sits below HKFRS. The dividend was funded from the AAMI cash, not from operations — whether this is the equipment-maker recovery cost or a quality break is what the H1 print will resolve.
The next six months hand the market three independent verdicts on the bull case.
- Late July / early August — H1 2026 interim print. First independent test of whether the +71.6% YoY bookings inflection converts into group gross margin above 40%, with 90+ past-due receivables trending below HK$250M. The single audited disclosure inside today's 80× P/E.
- Q3-Q4 2026 — SMT Strategic Options outcome. Formal review of the 46%-of-revenue placement segment, announced January 21, 2026. A sale or spin re-prices SEMI on the BESI/Hanmi pure-play peer set; 'retain' or 'partnership' perpetuates the conglomerate discount the bull case underwrites away.
- 3-6 months — new Group CEO. Robin Ng departed the May 7 AGM with no internal successor named. The next CEO's first 100 days set R&D intensity, the SMT decision, and capital allocation through the most expensive part of the hybrid-bonding race.
The 'broad-based portfolio is a hedge' story was retired in 2025.
Before: From the 2022 rebrand through 2024, management's signature defence was the SEMI+SMT 'unique broad-based portfolio' — diversification as a cushion through cycles. Capital was spread across SEMI back-end, SMT placement, the AAMI materials joint venture, the NEXX wafer-level plating unit, and AEC mainstream bonders in Shenzhen.
Pivot: In late 2025 the framing flipped. AAMI was sold to Zhizheng for HK$862M cash plus a HK$1,113M one-off gain (November 2025). The Shenzhen AEC plant was liquidated (August 2025). On January 21, 2026 the board opened a formal Strategic Options Assessment of SMT itself. NEXX was sold to Applied Materials for US$120M (May 2026).
Today: ASMPT is being remade as a focused advanced-packaging franchise. Three businesses sold or liquidated inside twelve months; the fourth — SMT — is the one that surfaces or buries the sum-of-parts. The CEO who built the broad-based pitch left at the May AGM; the new CEO will decide whether the reset finishes or stalls.
Watchlist. The moat is real but the multiple already prices the re-rate.
- For — TCB qualification at the centre of the AI cycle. Sole supplier C2S at TSMC CoWoS plus multi-customer HBM4 12H POR; TCB revenue +146% in FY2025; 15 additional C2S tool orders disclosed December 2025; Q1 2026 net profit +204% YoY.
- For — SMT separation could force a sum-of-parts re-rate. A clean sale or spin re-bases the SEMI peer set from broad-line (KLIC 6.8× EV/sales) toward BESI/Hanmi (35–50×). On a HK$72B market cap that gap is the upside the bull underwrites.
- Against — FY2025 quality is too soft for an 80× trough. 92% of pre-tax profit from a one-off JV gain; FCF -HK$139M; 90+ past-due receivables +69% YoY to HK$330M; dividend funded from an asset sale, not operations.
- Against — BESI owns the next node. 150+ hybrid bonders to 18 customers and first integrated production lines installed, versus ASMPT's gen-2 LITHOBOLT with no customer count disclosed. The HBM5 transition (2028–29) is where the moat either renews or unwinds.
Watchlist to re-rate: H1 2026 group gross margin (>40% confirms leverage, <38% breaks it); the language of the SMT outcome ('sale/spin' vs 'retain/partnership'); the new CEO's first capital-markets day and R&D-intensity stance.