ASMPT: Record Bookings and a Real TCB Inflection, but at 135.10 HKD There Is No Conservative-Case Margin of Safety

The HBM opportunity is less mature than the logic orders. A key memory customer was sampling ASMPT’s flux-based TCB process and qualifying the AOR fluxless process for 16-high HBM4. Qualification is commercially meaningful because semiconductor manufacturers rarely change a critical bonding process casually after high-volume approval. It is still a step before volume production. HBM4 product schedules, yields, stack architecture and the memory customer’s vendor-allocation decision determine whether sampling becomes a material revenue stream.

Competition prevents the TCB opportunity from being treated as an uncontested franchise. Hanmi Semiconductor is deeply entrenched in HBM thermo-compression bonding and markets its latest tools for HBM4. Besi supplies both advanced TCB and hybrid-bonding systems and reported first-half 2026 bookings growth of 116.5%, including strength in hybrid bonding. Other assembly-equipment manufacturers and captive systems also participate. ASMPT’s position appears strongest where its installed process knowledge, AOR approach and ability to serve several bonding architectures solve a customer-specific production problem. Its position is weaker if customers standardize around a rival’s HBM platform or move rapidly from TCB to hybrid bonding.

Hybrid bonding is both an opportunity and a threat. It enables copper-to-copper interconnection at pitches below the practical range of solder-bump-based TCB, with potential density and power advantages. Besi is the most visible listed beneficiary and already has production-oriented hybrid-bonding exposure. ASMPT’s second-generation hybrid-bonding platform had reached customer sampling by June 2026, which keeps it in the race but does not establish high-volume incumbency. Broad HBM substitution around 2029–2030 remains plausible, while selected logic applications are likely to adopt earlier. The transition will probably be application-specific rather than a single industry-wide cutover date.

The SMT strategic question has moved beyond media speculation but remains unresolved. ASMPT formally announced a strategic-options assessment on 21 January 2026. The company explicitly listed a divestiture, joint venture, spin-off and public listing, or retention and continued support as possible outcomes. At the 29 July interim results, SMT remained a continuing operation, and no completed transaction or selected option was announced. The correct status is therefore: a company-confirmed review, with no disclosed outcome as of the research date. A base valuation should continue to include SMT and should not award a pure-play advanced-packaging multiple to the entire group.

The ownership record also requires correction. Primary disclosures show that ASM International reduced its interest in stages before 2025: a roughly 12-percentage-point placement occurred in 2013, and sales during 2017 ultimately reduced its holding to approximately 25%. ASM International still held about 25% at the end of 2025. The asserted 2025 reduction from roughly 37% to 25% is not supported by the primary record. A future placement remains a legitimate technical overhang because a 25% block is large, but it should not be presented as the continuation of a 2025 sale. ASM International and ASMPT are separate listed companies operating at different points in the semiconductor chain.

ASMPT’s balance sheet buys it time to work through these choices. Cash and bank deposits were HK$5.88 billion at June 2026, net cash was HK$3.63 billion, and debt-to-equity was 0.13. Inventories rose to HK$7.25 billion from HK$6.30 billion in six months, while trade and other receivables increased to HK$5.43 billion from HK$4.24 billion. Customer advances rose even faster, to HK$1.95 billion, supporting the view that part of the working-capital build accompanies real orders. The competing interpretation is that the company is carrying substantial inventory and receivables into an unusually strong order cycle. Cash conversion will decide between those readings.

The market has already assigned substantial value to the AI packaging thesis. At HK$135.10, ASMPT was about 45% below its 52-week high of HK$244.40 but 108% above its HK$64.95 low. The share price fell 13.7% on 28 July, before the interim announcement, and declined again after results despite the strong headline figures. That reaction suggests expectations had risen faster than reported earnings, although event timing prevents attributing the entire decline to the results. Besi’s shares had also more than doubled during 2026 before its investor day, evidence that a sector-wide advanced-packaging re-rating accompanied ASMPT’s company-specific order momentum.

Trailing P/E is a poor standalone guide because ASMPT’s earnings contain discontinued-business losses, disposal gains, restructuring costs and a sharp cycle recovery. Market-data providers placed the trailing ratio near 39 times after the July correction, versus 8–10 times at the 2021–2022 cyclical peak and more than 40 times during the 2023–2024 trough. On a forward owner-earnings basis, this report estimates a ratio in the low-to-mid twenties, depending on backlog conversion and working capital. That is below the most aggressively valued advanced-packaging peers but above ASMPT’s old-cycle valuation center.

The qualitative portrait is company in transition. ASMPT is transitioning from a diversified back-end and electronics-assembly supplier valued mainly on semiconductor cycles into a portfolio whose most valuable growth assets sit in advanced packaging. The transition is operationally real: TCB orders, logic-customer adoption, photonics growth and SEMI margins provide evidence. It is incomplete: SMT remains almost half the business, HBM4 qualification has not yet become disclosed volume production, hybrid bonding is still at sampling, and cash conversion lags accounting recovery.

Vertical history and financial record

ASMPT began in 1975 as the Asian marketing and service arm of ASM International. The institutional logic was straightforward. Semiconductor manufacturing was shifting toward Asia, but equipment suppliers needed local sales, service and engineering support. Arthur del Prado’s ASM International supplied the parent-company context, while Patrick Lam built the Asian operation and pushed it beyond distribution. The business gradually added tooling, lead-frame-related capabilities, wire-bonding equipment and its own automated assembly systems. What began as a regional commercial arm became a manufacturer because local customer proximity exposed production problems that could be solved with purpose-built equipment.

The listing in Hong Kong in 1989 institutionalized that independence. The initial offering comprised 90 million new shares at HK$1.61 each, implying gross proceeds of about HK$144.9 million before expenses and a stated P/E of roughly 7.5 times. The capital-market story was an Asian semiconductor-equipment manufacturer positioned close to the region where assembly capacity was expanding. The low single-digit-to-high single-digit valuation reflected a cyclical industrial supplier rather than a scarce technology platform.

The first durable stage ran from the 1990s through the late 2000s. ASMPT expanded manufacturing and engineering in Singapore, Hong Kong, Malaysia and mainland China, while building positions in wire bonding, die bonding, encapsulation and lead-frame tooling. It became the largest supplier of semiconductor assembly and packaging equipment by revenue in 2002 according to its corporate history. Scale mattered because customers needed global support, spare parts and process engineering, while broad product coverage allowed the company to sell across several assembly steps. The lasting capability from this period was the ability to industrialize precision equipment in Asia and support it globally.

The second stage began with the 2011 acquisition of Siemens’ SIPLACE placement-equipment business and accelerated with the 2014 acquisition of DEK’s printing business. These transactions created the modern SMT segment. SIPLACE brought high-speed component placement; DEK added solder-paste printing and process-control exposure. The strategic rationale was diversification beyond semiconductor packaging into electronics manufacturing, along with a wider installed base and service network. The price was structural complexity: SMT customers, competitive dynamics and margins differ from those of semiconductor packaging.

ASMPT subsequently used smaller acquisitions to enter higher-growth niches. The 2018 purchase of AMICRA expanded high-precision die-attach capability for photonics and advanced packaging. Critical Manufacturing added manufacturing-execution software. The portfolio became unusually broad: traditional wire and die bonding, advanced packaging, photonics, board assembly and factory software. Breadth created cross-cycle resilience and customer access, but it also made the company harder to value and manage.

The 2020–2022 period exposed both sides of the model. ASMPT deconsolidated its materials business into the AAMI joint venture at the end of 2020, reducing direct exposure to lead frames and related materials. The subsequent semiconductor and electronics boom produced record 2021 revenue of HK$21.95 billion, record bookings of HK$26.12 billion and a HK$10.06 billion backlog. Continuing-operations EPS reached HK$7.72. The market could then treat ASMPT as a cyclical beneficiary with substantial operating leverage.

The cycle reversed sharply. Revenue fell from HK$21.95 billion in 2021 to HK$12.48 billion in 2024, a 43% contraction. Continuing-operations profit attributable to owners fell from HK$3.17 billion to HK$294 million. Consumer electronics, automotive and industrial customers worked through excess inventories, while mainstream assembly-equipment utilization weakened. Advanced packaging continued to develop, but it was initially too small to offset the group downturn.

The market briefly overlaid an event-driven narrative on the cyclical trough. In October 2024, ASMPT confirmed a preliminary, non-binding approach after reports identified KKR as the potential bidder. The shares initially rose sharply, but takeover discussions ended in November. The termination closed the formal offer period, and the shares fell more than 20% around the announcement. That episode did not alter the operating franchise. It did expose a valuation gap large enough to attract private-capital interest and showed how quickly event premium could disappear. No current take-private thesis should be included without a new formal disclosure.

The present stage began during 2024–2025 as TCB moved from customer engagement into record bookings and revenue. Management simultaneously simplified the portfolio. In 2025, ASMPT sold its remaining AAMI interest, recognizing a HK$1.11 billion disposal gain. In June 2026 it completed the US$120 million sale of NEXX, a deposition-equipment subsidiary that was loss-making in the first half. The SMT strategic review pushes the simplification question much further because SMT is a large, profitable segment rather than a small peripheral unit.

The company’s five-year financial record shows a deep cycle rather than steady compounding.

† 2024–2025 presentation reflects the NEXX discontinued-operation classification where applicable. ‡ Combined continuing and discontinued operations, to maintain consistency with consolidated cash flow.

Source: ASMPT annual reports and investor financial information; ratios are report calculations.

Revenue movement was driven primarily by equipment volume and product mix, not recurring subscription growth or sustained price increases. The 2021 peak reflected broad semiconductor, automotive, industrial and consumer demand. The 2022–2024 decline followed customer inventory correction and weaker capital expenditure. Recovery in 2025 came from SEMI, particularly TCB, while SMT revenue remained 1% below 2024 despite stronger late-year orders.

The gross-margin record reveals why mix matters. Group gross margin held near 40% through much of the downturn but fell to 37.8% in 2025 as both segments faced weaker mix and utilization. In the first half of 2026, higher TCB, high-end die-bonder and photonics volume lifted adjusted group gross margin to 41.2%. SEMI’s adjusted gross margin reached 46.5%, compared with 34.2% for SMT. The group therefore gains margin when advanced packaging grows faster than mainstream SMT, even before any asset separation.

Accounting earnings and cash flow match reasonably over a full cycle but diverge sharply in individual years. Aggregate operating cash flow for 2021–2025 was HK$9.10 billion, equal to 1.17 times aggregate attributable profit of HK$7.75 billion. After HK$1.85 billion of PP&E purchases, the ratio falls to 0.93. This is acceptable through-cycle conversion for an equipment manufacturer, although 2025 was poor: operating cash flow was only HK$243 million because receivables and inventory absorbed cash and restructuring provisions were paid.

Capital expenditure is modest relative to revenue because ASMPT’s economic investment is carried largely through expensed research and development rather than new wafer-fabrication plants. PP&E purchases averaged about HK$370 million over five years. Reported depreciation of PP&E was HK$339 million in 2025, close to the HK$382 million cash purchase figure, suggesting that most physical capex is maintenance or replacement. Capitalized intangible additions were a further HK$100 million in 2025. A defensible normalized estimate is HK$300 million–HK$350 million of annual maintenance capex and HK$50 million–HK$150 million of discretionary growth or expansion capex. The company does not disclose this split, so it remains an analytical estimate rather than a reported figure.

R&D is the larger economic commitment. ASMPT spent HK$1.93 billion on R&D in 2025, about 14% of revenue, with HK$1.16 billion allocated to SEMI and HK$767 million to SMT. Spending stayed almost flat from 2024 despite a weak revenue base, depressing short-term margin but preserving product development through the cycle. The first-half recovery shows the operating leverage available when volume returns against this largely fixed engineering base.

Balance-sheet risk is limited. June 2026 net cash of HK$3.63 billion equaled about HK$8.65 per share. Current assets exceeded current liabilities by HK$12.57 billion. Bank borrowings were HK$2.25 billion and unsecured, while available banking facilities were HK$3.02 billion. Goodwill and acquired intangibles require monitoring because AMICRA, Critical Manufacturing and other acquired operations carry execution risk, but leverage does not threaten solvency under a normal semiconductor downturn.

Returns on capital are cyclically volatile. Approximate attributable ROE exceeded 20% near the 2021 peak, fell into the low single digits in 2024 and recovered toward a low-double-digit annualized rate on first-half 2026 adjusted earnings. That pattern does not support a structural “compounder” label. The company earns attractive returns when capacity utilization, advanced-product mix and semiconductor capital spending align; it has yet to show that the new TCB mix can sustain high returns across a complete cycle.

ASMPT’s valuation history follows this cycle. At the 2021–2022 earnings peak, year-end P/E fell to about 8–10 times because investors anticipated normalization. During 2023–2024, reported P/E rose above 40 times as earnings collapsed, even though the share price was far below prior peaks. The current trailing ratio near 39 times therefore does not mean the shares are cheaper than in 2021. It means current earnings remain below the market’s expected run rate.

The 12-month price path contains both sector and company-specific components. The 52-week range of HK$64.95–HK$244.40 captures a rapid AI and advanced-packaging re-rating followed by a severe correction. Record TCB disclosures, rising bookings and peer enthusiasm drove the upward leg. The pre-results July selloff and weak reaction to strong interim figures indicate that investors had begun discounting more than a simple cyclical rebound. The stock at HK$135.10 still prices a sizable earnings recovery, but it no longer prices uninterrupted execution.

Business model, moat, governance, and industry cycle

ASMPT’s economic engine has three layers. Mainstream SEMI tools such as wire and die bonders generate volume through general semiconductor assembly cycles. Advanced-packaging products, including TCB and photonics tools, generate higher growth and potentially better margins but are more customer- and program-specific. SMT provides exposure to electronics assembly, AI-server boards, automotive electronics and industrial production. Software, services and spare parts support the installed base, although ASMPT does not disclose a sufficiently large recurring-revenue stream to value it like a software or service company.

The latest segment economics illustrate the divergence.

§ Group adjusted operating profit divided by group revenue.

Source: ASMPT 2026 interim results; group margin is a report calculation.

SEMI is the better-margin business, but SMT’s first-half acceleration was stronger in both absolute bookings and revenue growth. This undermines the simple claim that SMT is a declining asset that should be sold at any price. Its 11% segment margin and record second-quarter bookings give management more negotiating leverage than it had during the 2023–2024 trough. They also increase the opportunity cost of a low-priced disposal.

The cost base contains significant operating leverage. Engineering, application support, software development, global sales infrastructure and factories must be maintained before an incremental tool ships. Material, contract manufacturing, freight and sales commissions vary more directly with volume. R&D alone was HK$1.93 billion in 2025, while total group operating expenses were HK$4.56 billion. During the downturn, revenue fell much faster than expenses; during the first-half 2026 recovery, adjusted operating profit grew more than four times as fast as revenue.

The first genuine moat is process qualification and customer co-development. A bonding tool must achieve placement accuracy, thermal control, throughput, yield and package reliability inside the customer’s process. Once a tool becomes a production tool of record, replacing it can require requalification, process tuning, technician retraining and yield risk. ASMPT disclosed that a global IDM had operated Firebird chip-to-wafer TCB systems in high-volume manufacturing since 2024 before placing repeat orders. That sequence is stronger evidence of switching costs than a laboratory specification.

The AOR process is the second moat. Conventional flux assists oxide removal during bonding but can leave residue, complicate cleaning and affect reliability at fine pitch. ASMPT’s plasma AOR approach performs oxide removal without conventional flux in the bonding sequence. The moat is meaningful where the customer validates the whole process, rather than merely buying a machine. Its durability remains unproven across multiple generations and customers because competing fluxless and hybrid approaches can solve the same problem differently.

Breadth comes third. ASMPT can address chip-to-substrate, chip-to-wafer, logic, memory, photonics and mainstream back-end applications, and it can connect chip packaging with board-level assembly. Breadth provides customer access and manufacturing scale. It can also dilute focus. Besi generates materially higher gross margins from a narrower assembly-equipment portfolio, while Hanmi’s concentration in HBM TCB allows it to optimize around a smaller set of memory customers. Breadth is therefore an advantage in account coverage and cycle resilience, not proof of superior economics.

Continuous engineering investment is the fourth. ASMPT maintained R&D near HK$1.9 billion through the downturn and operates a global R&D organization of more than 2,200 people. This commitment supports TCB, hybrid bonding, photonics, high-end die bonding and software development. The spending becomes a moat only when it creates qualified production tools. NEXX’s losses and subsequent sale show that R&D breadth can also destroy value where commercial scale does not follow.

ASMPT lacks several classic moats. There is no network effect, scarce regulatory license or commodity cost advantage. Customers are technically sophisticated and large enough to negotiate price, split orders and sponsor competing vendors. Equipment revenue is transactional and cyclical. The installed base creates service and qualification benefits, but not an annuity large enough to neutralize capital-spending cycles.

Robin Ng became group CEO in May 2020 after serving as CFO and has more than 30 years of finance, audit and accounting experience. Katie Xu is group CFO, and Guenter Lauber is chief strategy officer and chairman of the SMT segment. The leadership mix combines financial discipline with long SMT operating experience, although the CEO’s background is less rooted in semiconductor process engineering than that of some equipment-company founders. John Lok Kam Chong was chairman at the research date.

Management’s capital-allocation record is mixed but improving. The SIPLACE and DEK acquisitions created a globally relevant SMT franchise rather than a stranded collection of assets. AMICRA helped establish photonics and high-precision die-attach exposure. Critical Manufacturing added software. Portfolio complexity increased all the same, and NEXX required further investment before being sold after losses. The AAMI and NEXX transactions, followed by the SMT review, show a more explicit willingness to exit assets outside the highest-priority back-end packaging areas.

Shareholder distributions have varied with the cycle. ASMPT paid HK$241.5 million of dividends in 2025 and declared a HK$0.97 first-half 2026 interim dividend, compared with HK$0.26 a year earlier. It did not repurchase shares during the first half of 2026. Share-based plans use revenue growth relative to peers and EBIT margin among performance conditions, which provides some alignment, although management’s direct ownership is modest relative to the 25% ASM International block.

ASM International’s remaining ownership can influence market liquidity and governance without making ASMPT a subsidiary. ASMPT’s accounts, management and business strategy are separate. The holding creates two opposing effects: an informed strategic shareholder may support discipline, while a future placement could place a large block into the market at a discount. Investors should monitor formal substantial-shareholder filings rather than infer sales from price weakness.

No evidence in the reviewed primary filings indicates accounting fraud or a material regulatory investigation. The accounts do contain recurring restructuring charges, acquisition-related impairments, foreign-exchange movements and non-HKFRS adjustments. In 2025, the HK$1.11 billion AAMI disposal gain was far larger than adjusted net profit of HK$467 million, so reported net income is unsuitable for measuring underlying earning power without reconciliation.

The relevant industry combines three cycles. Mainstream wire bonding, die bonding and SMT follow semiconductor and electronics inventory cycles. TCB and photonics follow a technology-adoption and AI-infrastructure cycle. Customer decisions remain part of a capital-expenditure cycle because tools are ordered in batches for new capacity and product ramps. ASMPT can therefore report structural advanced-packaging growth while its other businesses remain cyclical or regionally weak.

Industry estimates vary by scope. Yole expects the TCB equipment segment to grow at roughly 11.6% annually and hybrid-bonding equipment at 21.1%, with the two technologies contributing substantially to back-end equipment growth through 2030. ASMPT’s own estimate places its addressable TCB market above US$1.6 billion by 2028. Company TAM estimates tend to use broader product definitions than third-party market segments, so the numbers should not be treated as directly comparable.

The profit pool is shifting toward processes that solve interconnect density, yield and thermal-management constraints. Traditional assembly tools still ship in larger volumes, but advanced bonding commands more engineering content per system and sits closer to the customer’s product architecture. Suppliers that become tools of record for leading logic or HBM programs can earn strong margins. Suppliers that lose qualification may see demand disappear abruptly because a limited number of customers control the highest-value programs.

The bargaining structure is demanding. Foundries, IDMs and large outsourced assembly companies buy multiple systems and can qualify more than one supplier. They possess detailed yield data that equipment vendors cannot replicate independently. Upstream precision-component and motion-control suppliers can create bottlenecks, although ASMPT’s scale gives it purchasing power. The most attractive economics accrue to vendors with differentiated process IP and an installed high-volume base, rather than to vendors offering mechanically similar tools.

China represented 42.2% of first-half 2026 group revenue, up from 38.1% a year earlier. The exposure is concentrated mainly in mainstream SEMI, SMT, electric vehicles and electronics assembly rather than only leading-edge AI packaging. China supports utilization and volume but creates geopolitical, local-competition and customer-credit exposure. Export controls aimed at advanced semiconductor manufacturing could expand to selected packaging capabilities or components, while Chinese industrial policy may support domestic equipment alternatives. ASMPT’s non-US domicile reduces some direct jurisdictional exposure but does not isolate a global supply chain from US, Japanese or European controls.

The company’s top five customers represented 19% of first-half revenue, which is moderate at group level. That figure understates possible concentration inside TCB because advanced-logic and HBM production is controlled by a small group of customers. ASMPT does not disclose TCB revenue by customer, the percentage of TCB bookings associated with its largest order, or customer-specific backlog. The absence of disclosure is a valuation constraint because the growth engine may be considerably more concentrated than the group statistic.

Horizontal competition and current fundamentals

ASMPT’s closest listed comparison is Besi, but the two companies have become different economic models. Besi narrowed itself around high-margin assembly tools and became the public-market reference for hybrid bonding. Its first-half 2026 revenue increased 48.8% to €434.7 million, bookings rose 116.5% to €562.6 million, and gross margin reached 64.7%. Its long-term target calls for €1.7 billion–€2.2 billion of revenue and a 45%–55% operating margin, without a stated attainment date. Customers and investors pay for a concentrated advanced-assembly platform with unusually high incremental margins.

Besi competes with ASMPT in advanced die attach, TCB and the migration toward hybrid bonding. It announced an order for five latest-generation TCB systems worth approximately US$20 million in 2025, implying high value per tool. Its hybrid-bonding collaboration and installed customer engagements make it the clearest technological substitute if fine-pitch logic and HBM move beyond solder-based TCB faster than ASMPT expects. Besi is also more exposed to a valuation reversal because its shares trade on expectations of large hybrid-bonding adoption and its trailing P/E was about 72 times at 30 July.

Hanmi Semiconductor represents the more direct HBM TCB challenge. Its product portfolio is narrower and closely tied to Korean memory manufacturers. The company markets its TC Bonder line as critical HBM equipment and introduced systems for HBM4. Second-quarter 2026 revenue increased 39.5% to approximately ₩251 billion, while operating margin reached 51.9%. The first quarter had been much weaker, illustrating the lumpiness of customer acceptance and shipment schedules. Customers choose Hanmi for HBM process history, local support and established memory-production qualification.

Claims that Hanmi controls more than 70% of the global HBM TCB market come from reports distributed or highlighted by the company and should be treated cautiously. The direction is credible, since Hanmi is a major HBM incumbent, but precise global share is hard to verify because captive systems, customer definitions and tool generations differ. ASMPT’s memory thesis therefore rests on taking a meaningful second-source or program-specific position rather than assuming an open market without an incumbent.

SÜSS MicroTec is an adjacent advanced-packaging peer rather than a direct general-purpose TCB competitor. It supplies lithography, temporary bonding and other wafer-processing systems used around advanced packaging and heterogeneous integration. First-quarter 2026 orders rose 69.5% to €149.3 million, while revenue fell to €86.5 million and EBIT margin was 4.3% as the company entered a transition year. Customers choose SÜSS when wafer-level process expertise and specialized coating, alignment or temporary-bonding steps matter. Its exposure provides a useful read on advanced-packaging capacity, but its product economics do not map directly onto ASMPT’s die-bonding franchise.

DISCO is further upstream in the package-formation process. It dominates precision cutting, grinding and polishing systems rather than TCB. Its fiscal first-quarter 2026 operating margin of 42.9% reflects strong process specialization, consumables and an installed base in wafer singulation and thinning. DISCO is a quality benchmark for what happens when an equipment supplier owns a critical process niche, but it is not a direct substitute for ASMPT’s bonders.

Kulicke & Soffa is the broader back-end comparison. It competes in wire bonding, advanced packaging and electronics assembly, with exposure to many of the same mainstream cycles. Its gross margins and balance sheet can be attractive, but trailing P/E near 80 times in July 2026 reflected depressed earnings and strong recovery expectations rather than a stable premium franchise. It is more relevant to ASMPT’s mainstream SEMI portfolio than to its highest-end TCB differentiation.

Ibiden belongs on the industry map as a high-end package-substrate supplier. It does not sell competing equipment, but its AI substrate capacity and customer investment plans influence demand for tools that assemble advanced packages. It is best used as a demand read-through rather than a valuation peer.

A narrow quantitative comparison reinforces the business-model differences. Market capitalizations have been translated at illustrative 30 July rates of HK$9.10 per euro and HK$0.0057 per Korean won; translation differences do not affect the relative conclusions.

§ Precise comparable net-cash figures were not standardized in this review.

Source: company results and market data as of late July 2026. Different reporting periods and adjusted definitions limit direct comparability.

Besi’s multiple reflects its narrower advanced-packaging identity and much higher gross margin. Hanmi trades on scarcity in HBM TCB and extreme recent profitability. SÜSS still gets an advanced-packaging premium despite weak current utilization. ASMPT’s lower multiple is partly justified by the lower-margin SMT business, mainstream SEMI exposure, Hong Kong listing discount and more complex portfolio. The discount could narrow if TCB becomes a larger share of earnings or SMT is separated on favorable terms. It could remain permanent if the market views ASMPT’s breadth as a source of cyclicality and lower returns.

ASMPT’s niche is best described as a diversified technology supplier becoming a leading challenger in advanced bonding. It is not the uncontested TCB leader in every market. Its foundry and outsourced-assembly engagements in chip-to-substrate logic look strong, and repeat IDM orders validate the chip-to-wafer logic platform. In HBM it is challenging an entrenched supplier set and remains in qualification for important future programs. In hybrid bonding it is a sampler rather than a disclosed volume-production incumbent.

Customers choose ASMPT when they value process flexibility, AOR capability, global support and the option to source several package-assembly technologies from one vendor. Besi wins on hybrid-bonding leadership and high-end die-attach specialization, Hanmi on established HBM production history and proximity to Korean memory ecosystems, SÜSS on specialized wafer-level preparation processes. Competition here is segmented, not a single winner-takes-all equipment category.

The last four reported quarters trace a clear inflection. Second-half 2025 revenue accelerated as TCB, photonics, AI-server and Chinese EV demand improved. Fourth-quarter group revenue rose 30.9% year on year, but gross margin remained only 35.8%, partly because of product mix and a cancellation-related inventory provision. First-quarter 2026 revenue rose 32% and bookings 71.6%. Second-quarter revenue rose 52.1% and bookings 97.6%, while adjusted gross margin improved to 42.5%. The business has moved from bookings recovery to revenue and margin conversion, although cash conversion remains incomplete.

Management has raised the operating trajectory faster than the market expected at the start of 2026. The first-quarter guidance midpoint was already above then-current consensus, actual first-quarter revenue reached US$507.9 million, and second-quarter revenue of US$630 million exceeded guidance. Third-quarter midpoint guidance of US$660 million implies further sequential growth. Management also expects high-single-digit sequential bookings growth, led by TCB and photonics.

SEMI’s second-quarter adjusted segment margin reached 23.7%, compared with 18.1% for the half, showing substantial incremental leverage when high-end products ship. SMT’s second-quarter margin reached 13.9%, benefiting from volume and mix. These quarterly margins should not be extrapolated mechanically: equipment acceptance timing and a few large programs can move revenue between quarters. The reasonable conclusion is that both segments can earn materially more than their 2025 margins at current volume, with SEMI retaining the superior structural gross margin.

The market is trading four linked expectations: that most of the bookings surplus becomes revenue within several quarters, that SEMI gross margin stays in the mid-to-high forties as TCB scales, that ASMPT wins meaningful HBM4 business rather than remaining a qualification supplier, and that any SMT strategic transaction improves rather than dilutes per-share value. Only the first two have substantial current financial evidence. The HBM and portfolio outcomes remain prospective.

The bull case argues that bookings are broad, customer validation is strengthening, and first-half operating leverage has only begun. More than 50 chip-to-substrate TCB tools, a repeat chip-to-wafer IDM order, leading-logic deliveries, photonics growth and HBM engagements create several growth vectors. The company’s estimated TCB addressable market above US$1.6 billion by 2028 is large relative to current advanced-packaging revenue.

The bear case begins with the same facts but interprets their durability differently. Orders are concentrated in expensive tools and may be lumpy. Group-level customer concentration does not reveal TCB concentration. A 2025 cancellation already required an inventory provision. HBM qualification may not convert, and a faster hybrid-bonding transition could shorten the TCB growth window. The share price’s rise from the 52-week low shows that these risks are no longer being priced against a depressed starting point.

The SMT review could produce a material valuation event, but its direction is ambiguous. A sale or listing at a strong multiple could reveal SEMI’s economics and fund investment or distributions. A low valuation, retained liabilities or stranded corporate costs could reduce earnings more than the proceeds compensate. Retention could be rational if SMT bookings convert and margins remain in double digits. Since the outcome is unannounced, a base case should assign no transaction premium.

Valuation, risks, catalysts, and tracking

Historical P/E analysis is distorted by the cycle. ASMPT traded near 8–10 times earnings at the 2021–2022 profit peak and above 40 times when earnings collapsed in 2023–2024. At HK$135.10, market-data estimates place trailing P/E around 39 times, below the trough multiple but far above old peak-cycle valuations. The proper comparison is normalized owner earnings and the expected mix of SEMI and SMT, rather than last year’s reported EPS.

Peer valuation offers context rather than proof of cheapness. Besi at roughly 72 times trailing earnings and Hanmi near 77 times embody aggressive expectations for hybrid bonding and HBM. SÜSS near 42 times also carries an advanced-packaging premium despite a transition-year earnings decline. The entire group has experienced an AI-related re-rating, so ASMPT’s discount does not by itself create a margin of safety.

The owner-earnings bridge begins with cash conversion. Five-year aggregate operating cash flow equaled 1.17 times attributable net profit, while operating cash flow after PP&E purchases equaled 0.93 times profit. The through-cycle relationship is acceptable. The 2025 figure was poor because working capital and restructuring absorbed cash. For valuation, this report deducts HK$300 million–HK$350 million of maintenance capex from normalized cash generation and does not count disposal proceeds as recurring earnings.

At the current price, 2025 owner earnings were effectively negligible because reported operating cash flow did not cover PP&E purchases. That result understates forward earning power because it coincided with the start of a large bookings ramp and restructuring payments. A normalized 2026 owner-earnings estimate of HK$2.0 billion–HK$2.4 billion implies a yield of approximately 3.5%–4.2%, equivalent to 24–29 times owner earnings. This is materially less extreme than the trailing P/E but still demands continued conversion.

The scenario valuation assumes SMT remains consolidated through the forecast period because no transaction outcome has been announced. Any transaction would require a separate proceeds, tax, stranded-cost and retained-liability analysis.

This is valuation-scenario analysis within a research framework, not investment advice. Inputs are report estimates based on company results, current net cash and 419.4 million shares.

The conservative scenario does not assume a collapse. It assumes order growth normalizes, 2026’s high book-to-bill produces only a partial 2027 revenue benefit, and group operating margin settles near the low double digits. A 22-times ex-cash multiple remains above old peak-cycle valuations because the mix now contains a larger advanced-packaging business. The resulting HK$108 value is below the market price, meaning the current price offers no protection against merely ordinary execution.

The base case assumes revenue returns above the 2021 peak by 2027 only modestly, but with a better mix: more TCB, photonics and high-end die bonding, accompanied by a profitable SMT recovery. A 14%–15% group operating margin is below the highest first-half SEMI margin and accounts for SMT dilution and corporate expense. The 24-times multiple requires evidence that advanced packaging has raised through-cycle quality.

The optimistic case requires HBM qualification to become volume business, logic TCB orders to expand beyond the currently disclosed customers, photonics growth to persist, and SMT bookings to convert without margin erosion. It also assumes hybrid bonding expands the addressable market rather than prematurely displacing ASMPT’s TCB installed base. The 27.5-times multiple is below current Besi and Hanmi trailing valuations but high for ASMPT’s historical profile.

The expectation gap is concentrated in bookings quality. The next results need to show more than another high number. Investors will look for third-quarter revenue near or above the US$660 million midpoint, sequential bookings growth as guided, SEMI gross margin at or above the mid-forties, and evidence that inventories and receivables are not consuming all incremental earnings. A disclosed HBM4 volume order or further chip-to-wafer repeat order would strengthen the strategic thesis. A lower book-to-bill caused by revenue conversion would be acceptable; a lower ratio caused by order weakness would not.

The base case’s most fragile assumption is the owner-earnings contribution from advanced packaging. If the incremental TCB and photonics contribution reaches only 70% of the modeled level, base owner earnings fall by roughly HK$0.3 billion–HK$0.4 billion. At the same multiple, base value falls from approximately HK$155 to about HK$135–HK$138, close to the current price. This sensitivity shows that the market does not require the full optimistic case, but it does require substantial conversion.

If owner earnings remain flat for three years and ASMPT exits the period at the same valuation, investor return would consist mainly of dividends. A normalized dividend yield of roughly 1.5%–2.0% would be below the Hong Kong 10-year government-bond yield of approximately 3.5% on 30 July 2026. Under that assumption, there is no margin of safety at this buy price.

Margin-of-safety sufficiency verdict: none. The company can still generate a positive return from current levels, but that return depends on earnings growth. The price is above the conservative value and roughly aligned with a reduced version of the base case. This is a good improving business at a price that requires continued improvement.

The first permanent-loss risk is TCB order and customer concentration: medium probability, high impact. The observable indicators are SEMI bookings, disclosed repeat orders, top-five customer concentration, cancellation provisions and inventory growth. If one leading logic or memory customer postpones a production ramp, ASMPT could lose both revenue and factory absorption while retaining customized inventory. The transmission path runs from bookings to lower revenue, lower SEMI gross margin, weaker cash conversion and multiple compression.

Technological substitution is the second risk. Probability is low-to-medium over 12 months but medium over five years; impact is high. The indicators to follow are hybrid-bonding production orders at Besi, ASMPT’s progression from sampling to customer qualification, interconnect-pitch roadmaps and whether HBM4 16-high systems use TCB, hybrid bonding or both. A faster transition would reduce the useful life of TCB R&D and installed-base advantages. A slower one would extend ASMPT’s TCB runway.

Third comes poor SMT strategic execution, at medium probability and medium-to-high impact. Investors should track formal announcements, transaction valuation, tax leakage, retained liabilities, corporate-cost allocation and employee/customer retention. Selling SMT cheaply could remove a recovering profit stream and leave fixed costs at the parent. Retention without a clear strategic rationale could preserve the conglomerate discount.

The fourth risk, working-capital deterioration, is medium on both probability and impact. At June 2026, inventory had risen 15% in six months and trade and other receivables 28%, while customer advances rose 79%. Advances support the order-quality case, but inventory and receivables need to convert into cash. An alert would be inventory and receivables continuing to grow materially faster than revenue for two quarters while operating cash flow remains weak.

Valuation compression is the fifth risk: medium-to-high probability, high impact, because advanced-packaging peers trade well above historical equipment multiples. A sector-wide move from 40–80 times trailing earnings toward 20–30 times could affect ASMPT even if its own estimates rise. The Hong Kong listing discount limits relative valuation, but it does not prevent absolute de-rating.

Sixth is shareholder overhang. Probability is medium and fundamental impact is low, though share-price impact could be medium. ASM International’s approximately 25% interest could be placed in one or more blocks, increasing free float but pressuring price. The trigger is a formal substantial-shareholder filing or placement announcement. The absence of a 2025 sale does not eliminate future disposal risk.

Positive catalysts include another quarter of book-to-bill above 1, HBM4 volume conversion, additional ultrafine-pitch chip-to-wafer orders, photonics revenue exceeding the first-half run rate, SEMI gross margin holding above 46%, and an SMT transaction at a valuation that exceeds the implied value embedded in ASMPT. A sustained improvement in operating cash flow would be particularly powerful because it would validate both backlog quality and owner earnings.

Negative catalysts include a TCB cancellation, loss of a logic or memory qualification, third-quarter revenue below guidance, SEMI gross margin below the mid-forties, inventory provisions, an unfavorable SMT transaction, an ASM International block placement, or evidence that a competitor’s hybrid-bonding system has entered HBM volume production earlier than expected.

ASMPT’s investor calendar identifies an October third-quarter results conference but had not posted a precise 2026 date in the retrieved calendar as of the report date. The company historically reported third-quarter results in the final week of October, so “late October 2026” is an expectation rather than a confirmed date.

The dashboard should be read as a chain. Bookings measure demand, gross margin reads product and utilization quality, and inventory, receivables and operating cash flow measure conversion. Customer disclosures measure concentration, competitor hybrid orders measure the technology window, and valuation determines how much execution the investor is already paying for.

Cross-synthesis and final research conclusion

Looking vertically, ASMPT has proven a capability more durable than any single product: it can move from regional support into precision-equipment design, scale manufacturing in Asia, integrate acquired technologies and maintain customer engineering across several semiconductor cycles. The company survived the transition from lead-frame-era assembly to automated wire and die bonding, expanded into SMT through large European acquisitions, and is now participating in chiplet, HBM and photonics packaging. The common thread is process industrialization close to Asian customers.

Past success came from both era tailwinds and management capability. The migration of semiconductor assembly to Asia gave the original business its opening, and the global electronics boom supported SMT expansion. The 2021 profit peak was heavily cyclical. Management still had to convert those tailwinds into products, service networks and acquisitions that retained customer relevance. The evidence does not support either extreme interpretation: ASMPT is neither a passive cycle passenger nor a business that compounds independently of its cycle.

Those historical capabilities remain present. R&D stayed near HK$1.9 billion through the downturn. The company converted chip-to-wafer TCB into repeat IDM orders, secured large chip-to-substrate orders and advanced AOR into ultrafine-pitch customer engagement. Photonics moved from a niche acquisition rationale into visible AI-infrastructure revenue. The first-half segment margin shows that the platform can earn attractive profits when mix and utilization align.

The horizontal comparison narrows the advantage. Besi has higher margins and a stronger disclosed hybrid-bonding position, Hanmi deeper HBM TCB incumbency, DISCO a more focused process monopoly. ASMPT’s advantage is the combination of TCB architectures, AOR, broad customer coverage, global support and portfolio breadth. Customers weighing several packaging routes value that combination. It has not yet produced peer-leading returns because the portfolio contains lower-margin SMT and mainstream assembly equipment.

ASMPT’s weakness is partly temporary and partly structural. Low utilization and weak mainstream demand were temporary; first-half 2026 margins show recovery. SMT’s lower gross margin is structural. So is TCB customer concentration, because leading-edge packaging has few buyers, and so is the hybrid-bonding threat, whatever its timing. Hong Kong’s valuation discount relative to scarce European and Korean advanced-packaging names may also persist.

The current valuation rewards more than past success but does not fully pre-spend the optimistic outcome. At HK$135.10, the shares are below the report’s HK$155 base value and far below the HK$228 optimistic value. They are above the HK$108 conservative value. The price therefore embeds meaningful backlog conversion and improved margins while leaving upside for HBM volume wins or a favorable SMT transaction.

The market’s most likely misjudgment is treating the debate as a binary choice between “TCB winner” and “hybrid-bonding loser.” Packaging transitions usually fragment by pitch, package type, cost, throughput and yield. TCB can continue growing in chip-to-substrate, selected chip-to-wafer and HBM applications while hybrid bonding captures the finest-pitch designs. ASMPT’s long-term outcome depends on earning money from both transition paths, not predicting one universal cutover date.

A second possible misjudgment concerns SMT. Market enthusiasm favors a sale because it would simplify the advanced-packaging story. First-half data show that SMT is currently a growing, double-digit-margin business with a large backlog. The right transaction price has risen with those results. An attractive separation could reveal value; a disposal justified only by narrative purity could destroy it.

Over the next year, the decisive variables are order conversion, SEMI gross margin, HBM qualification and cash flow. The company needs to convert its implied backlog without accumulating customized inventory or extended receivables. A HBM4 volume order would broaden the TCB thesis beyond logic. Third-quarter guidance and bookings provide the first test.

Over three years, customer breadth and technology migration matter more. ASMPT must turn several large TCB engagements into a repeatable franchise, move hybrid bonding from sampling into production qualification, and decide whether SMT belongs inside the group. The best outcome is a higher-quality SEMI business with several logic, memory and photonics customers, plus either a well-priced SMT separation or a demonstrably profitable retained segment.

Over five years, the question becomes whether ASMPT owns a durable advanced-packaging process position after the current TCB wave. A company still dependent on one generation of solder-based bonding would deserve a cyclical multiple. A company supplying qualified TCB, hybrid bonding, photonics and high-precision die attach across multiple architectures could sustain a structurally higher return on capital.

The company becomes a better investment if three conditions align: the price offers protection against the conservative case, SEMI margins remain above 45% while operating cash flow follows profit, and customer disclosures show repeat orders across more than one logic and memory program. It becomes a weaker investment if bookings slow before revenue converts, hybrid-bonding competitors enter HBM volume production while ASMPT remains at sampling, or SMT is sold on terms that dilute normalized earnings.

Core bull reasons:

• First-half bookings increased 85.1% to HK$12.75 billion and book-to-bill reached 1.43, providing greater forward revenue visibility than at any point since the 2021 upcycle.
• SEMI adjusted gross margin reached 46.5% and segment margin 18.1%, showing that advanced-product mix can raise group economics as TCB and photonics scale.
• More than 50 chip-to-substrate TCB orders, a repeat chip-to-wafer IDM order and ultrafine-pitch logic deliveries validate several production routes rather than one laboratory program.
• Net cash of HK$3.63 billion provides financial capacity to fund R&D and absorb a cyclical setback without equity financing.
• A favorable SMT transaction could expose SEMI’s superior margins and reduce the conglomerate discount, while retention still offers earnings value if SMT’s record bookings convert.

Core bear reasons:

• HBM4 16-high remains in sampling and qualification, so the most valuable memory opportunity has not become disclosed volume revenue.
• Competitors possess stronger positions in key future paths: Hanmi in HBM TCB and Besi in hybrid bonding, while ASMPT’s second-generation hybrid tool remains at customer sampling.
• Operating cash flow fell to HK$243 million in 2025 despite reported profit, and first-half 2026 inventory and receivables rose materially, leaving backlog quality unproven in cash terms.
• The current price is above the HK$108 conservative value and would produce a return below the Hong Kong 10-year bond yield if owner earnings remained flat for three years.
• A large shareholder still holds approximately 25%, creating a block-placement overhang, while the widely repeated claim of a 2025 sell-down is factually incorrect.

The first pre-mortem script begins in 2027. A leading logic customer qualifies Besi’s hybrid-bonding platform for a package that ASMPT had expected to serve with ultrafine-pitch TCB, while a memory customer retains Hanmi as its primary HBM4 supplier. ASMPT receives fewer repeat orders, one customized order is postponed, and inventory provisions return. SEMI gross margin falls from 46.5% toward 40%, group operating margin drops from approximately 14% to 8%–9%, and normalized EPS falls toward HK$4. The market cuts the ex-cash multiple from 24 times to 15 times. Including net cash, the share price could fall into the HK$65–HK$75 range, a loss of roughly 45%–52% from HK$135.10.

The second script centers on portfolio execution. ASMPT sells SMT during 2027 at a low cyclical multiple but retains corporate, pension, lease and separation costs. SMT customers delay orders during ownership transition, and the proceeds are used partly for restructuring rather than a large distribution. At the same time, ASM International places part of its remaining stake. Continuing owner earnings settle near HK$1.8 billion and the market applies 17 times because SEMI still includes mainstream cyclical tools rather than constituting a pure advanced-packaging company. The resulting value could be HK$80–HK$90 per share.

ASMPT is a technically credible advanced-packaging contender embedded in a broader cyclical equipment group. The bookings inflection, logic-customer validation and margin recovery are real. The market still requires evidence that HBM qualification, cash conversion and hybrid-bonding development can turn the current order wave into durable owner earnings.

At HK$135.10, the shares sit near the low end of the report’s acceptable-hold range. They are not priced for the optimistic case, but they are priced above a conservative outcome in which bookings normalize and group margins remain merely adequate. The most important concern is the possibility that concentrated TCB orders and a temporary technology window are capitalized at a durable-growth multiple before cash and customer diversification prove the thesis.

A lower entry price would improve the asymmetry. A higher research conviction would require repeat HBM or logic production orders, SEMI gross margin remaining above 45%, and a visible improvement in operating cash flow. A favorable SMT decision would be additive but is not required in the base case.

【Company-profile scores】

• Fundamental quality: medium
• Growth: high
• Moat: medium
• Financial soundness: strong
• Management credibility: medium
• Valuation attractiveness: medium
• Risk level: medium
• Suitable investor type: cyclical and long-term growth investors able to monitor technology qualification and order conversion

【Investment rating】

• Rating: Hold
• One-line thesis: Record bookings and improving SEMI margins support earnings growth, but current valuation still requires substantial TCB conversion and offers no conservative-case margin of safety.
• Ideal buy price:

【Ideal Buy Price】80–88 HKD

Basis: at least 20% below the HK$108 conservative value, providing protection against slower booking conversion and a return toward normalized equipment multiples.

• Acceptable hold price: HK$132–HK$178, approximately within 15% of the HK$155 base value
• Clearly overvalued price: HK$253–HK$285, beginning at more than 10% above the HK$228 optimistic value
• Current-price classification: acceptable hold
• Whether to wait for a better price: yes for new capital; the preferred trigger is HK$88 or below while group book-to-bill remains above 1.1, SEMI gross margin remains at least 45%, and no major TCB customer loss is disclosed. The opportunity cost is missing continued order conversion or a favorable SMT transaction.
• Target holding horizon: 3–5 years
• Expected annualized return: approximately -6% in the conservative scenario, 6% in the base scenario and 21% in the optimistic scenario over three years, including an assumed 1.5% annual cash yield
• Max-loss risk: approximately 45%–55% if a major logic or HBM qualification is lost, SEMI gross margin falls toward 40%, and the ex-cash multiple compresses to about 15 times
• Reassessment-trigger signals:
  • Group book-to-bill below 1.0 for two consecutive quarters before backlog converts
  • SEMI adjusted gross margin below 43% for two consecutive quarters
  • Inventory and receivables growing more than 20 percentage points faster than revenue while operating cash flow remains weak
  • A disclosed loss or cancellation of a leading logic or HBM TCB program
  • Hybrid-bonding competitors entering material HBM volume production while ASMPT remains in sampling
  • An SMT transaction whose proceeds, stranded costs and lost earnings imply less value than retention

【Valuation Range】

• current: 135.10 HKD (close as of 2026-07-30)
• bear (conservative · ideal buy zone): [80, 88]
• base (fair · acceptable hold zone): [132, 178]
• bull (optimistic · above the clearly-overvalued line): [253, 285]

Sources and research uncertainties

The primary evidence base consists of ASMPT’s 2025 annual report, 2026 first-quarter announcement, 2026 interim results, strategic-options announcement, NEXX disposal filing, corporate history, investor calendar and technology releases. The annual report supplies audited financial statements, segment data, cash flow, R&D spending, ownership disclosures and the formal SMT review description. The interim filing supplies the latest bookings, margins, balance sheet and guidance.

Competitor evidence comes principally from Besi, SÜSS MicroTec, DISCO and Hanmi corporate releases and product materials, supplemented by Reuters for market reactions and industry context. Yole data provide directional market-growth estimates. Market prices and valuation ratios come from dated market-data pages and should be treated as point-in-time measures.

The first uncertainty is TCB customer concentration. ASMPT reports group top-five customer exposure but does not disclose the largest TCB customer, customer-level bookings, or the percentage of advanced-packaging revenue tied to individual logic and memory programs. Group diversification may therefore overstate diversification inside the growth engine.

The second is backlog quality. The company reports bookings and selected cancellation effects but does not disclose standard cancellation provisions, customer deposits by segment, average delivery lag, backlog gross margin or expected quarterly conversion. The implied HK$10 billion first-half backlog bridge is a calculation, not a company-reported ending balance.

Third comes the maintenance-versus-growth capex split. ASMPT reports total PP&E and intangible additions but does not classify maintenance capex. The HK$300 million–HK$350 million maintenance estimate is based on depreciation, capex stability and the asset-light nature of equipment assembly.

The fourth uncertainty is SMT transaction value. No selected option, buyer, timetable or valuation had been announced by 31 July 2026. The report therefore assumes retention and gives no transaction premium. A transaction could materially alter revenue, overhead, cash, peer selection and the valuation range.

Fifth is hybrid-bonding timing. Public roadmaps differ by supplier and application. Logic, HBM stack height, pitch, throughput and yield may produce several adoption curves rather than one 2029–2030 inflection. The valuation treats broad substitution timing as uncertain and requires ASMPT to progress beyond customer sampling.

Other tickers mentioned

• BESI.AMS: closest listed advanced die-attach, TCB and hybrid-bonding valuation peer
• 042700.KRX: Hanmi Semiconductor, a major incumbent in HBM thermo-compression bonding
• SMHN.ETR: SÜSS MicroTec, an adjacent wafer-level and advanced-packaging equipment supplier
• 6146.TSE: DISCO, a process-quality benchmark in wafer cutting, grinding and polishing
• KLIC.US: Kulicke & Soffa, a broader back-end and wire-bonding competitor
• 4062.TSE: Ibiden, a package-substrate supplier used as an AI advanced-packaging demand read-through
• ASM.AMS: ASM International, separate front-end equipment company and ASMPT’s approximately 25% shareholder
• AMAT.US: Applied Materials, a strategic investor and hybrid-bonding partner associated with Besi
• LRCX.US: Lam Research, cited in reporting on strategic interest in advanced-packaging assets
• 603991.SHG: listed buyer involved in ASMPT’s disposal of its AAMI interest

This report is based on public information and does not constitute investment advice. Markets carry risk; invest with caution.