At the close on June 12, 2026, BE Semiconductor Industries (BESI.AS, hereafter BESI) traded at EUR 317, for a market capitalization of about EUR 24.38 billion. The stock is up 167.6% over one year, against roughly 45.9% for the AEX index (StockAnalysis). The median 12-month target across 23 analysts is EUR 252.55 — the market price sits about 25.5% above the sell side’s midpoint. The same company reported fiscal 2025 revenue of EUR 591.3 million, down 2.7% year over year. Its trailing price-to-earnings ratio stands at roughly 148 to 161x, more than five times its 10-year median of about 27x.
A company with falling revenue carrying a historic valuation: what fills the gap is not current earnings but two options. The first is a technology-succession option. Hybrid bonding is the next generation after TC (thermo-compression) bonding, and BESI is the category’s acknowledged leader, with the most accurate tool on public record and the earliest volume qualification. First-quarter 2026 total orders reached EUR 269.7 million, up 104.5% year over year; unit orders more than doubled quarter over quarter; the customer count expanded to 20. But volume production on the HBM side does not arrive until after 2027. The second is a takeover option. Applied Materials (AMAT) bought a 9% stake in April 2025; in March 2026, media reported that both Lam Research and AMAT were weighing a bid, and that BESI had retained Morgan Stanley. The rumor has not become an offer.
This piece does one thing: it peels EUR 317 into three layers — the cash flow of the mainstream cycle, the option premium on hybrid bonding, and the possibility of being acquired — and then asks where the floor is once all three are separated.
Lay the timeline straight.
April 14, 2025: Applied Materials announced the purchase of a 9% stake in BESI as a strategic investor, becoming the largest outside shareholder, while stating it would not seek a board seat and did not plan to increase the position (Applied Materials IR). The stake rests on five years of groundwork: the two companies have jointly developed a fully integrated die-based hybrid bonding system since 2020.
February 19, 2026: BESI reported fiscal 2025 results. Full-year revenue of EUR 591.3 million, down 2.7% — but fourth-quarter orders of EUR 250.4 million, up 105.4% year over year and 43.3% quarter over quarter, driven by hybrid bonding and 2.5D demand from Asian OSATs (company IR). Revenue shrinking while orders explode: the two-speed structure became impossible to ignore with this report.
March 13, 2026: multiple outlets reported that both Lam Research and Applied Materials were considering an acquisition of BESI and that the company had retained Morgan Stanley to evaluate potential offers. The stock rose 10.2% that day, closing around EUR 194 (Investing.com, IndexBox; all on a reported-rumor basis). As of this writing there is neither a formal offer nor a formal denial — the rumor has hung in the air for three full months.
April 23, 2026: first-quarter results. Revenue of EUR 184.9 million, up 28.3% year over year; orders of EUR 269.7 million, up 104.5% year over year and another 7.7% quarter over quarter; hybrid bonding unit orders more than doubled sequentially, clearing the prior peak of the second quarter of 2024; customers up from 18 to 20; backlog of EUR 268.7 million, more than doubling within the quarter (company IR). Second-quarter guidance: revenue up 30% to 40% sequentially (roughly EUR 240 to 260 million), gross margin of 64% to 66%. The stock rose 3% that day.
June 12, 2026: the close at EUR 317. Measured from the rumor-day close of about EUR 194, that is a gain of roughly 63% in three months (a rough calculation across two closing prints) — a stretch in which the order surge and the takeover expectation are entangled. The later sections take them apart.
The qualification economics of TC bonders is something we took apart in Hanmi Semiconductor: The Results Broke, the Narrative Didn’t: the back-end bonding equipment business is not about selling machines but about getting into the production line. Tools must pass customer qualification, qualification is tied to yield learning, and yield learning is tied to the customer’s ramp schedule — switching suppliers means rerunning the entire learning curve. In the TC generation, those economics gave Hanmi Semiconductor a 46% operating margin in its best year, and also left it watching its share get split in half almost overnight when the generation turned.
Hybrid bonding is the next station on the same line. TC bonding joins stacked dies through microbumps under heat and pressure; hybrid bonding removes the bumps entirely, connecting copper directly to copper — higher interconnect density, lower power, and a better long-run cost path. The price is engineering difficulty: copper-to-copper alignment tolerance must stay within 1 micron, and that precision threshold is the core of BESI’s moat. TSMC’s SoIC process has already qualified BESI tools in volume production; the fully integrated system co-developed with AMAT over five years is a second layer of differentiation. Reuters has described BESI’s hybrid bonding tool as the world’s most accurate; industry analyses generally rank it first in the category, though no official share statistics exist.
The order book is the hardest evidence behind this option — but cut the attribution first: total orders doubling year over year for two consecutive quarters cannot be booked wholesale to this layer — on the company’s own account, fourth-quarter orders were driven in part by 2.5D demand from Asian OSATs, and the first quarter’s 28.3% revenue growth came mainly from high-end mobile and 2.5D AI, which belong to the mainstream layer’s conventional business. Strip those out, and what lands specifically under hybrid bonding is this: unit orders more than doubled sequentially in a single quarter, clearing the prior peak of the second quarter of 2024, and customers grew from 18 to 20, spanning four application domains — logic, memory, photonics, and mobile; on the memory side, two evaluation tools have shipped to a second memory customer, with three memory customers now in evaluation or deployment (company IR, earnings-call summaries). The company is scaling annualized capacity from 180 units to 250, and building a new line in Vietnam to absorb mainstream assembly so Malaysian capacity can focus on wafer-level advanced packaging — capacity moves that corroborate the demand as more than a one-off.
But the cadence has to go on the table too. On TrendForce’s account: in 2026, hybrid bonding on the HBM side remains in development and pilot production — a qualification year; mainstream volume production arrives after 2027; by the HBM4E era (around 2028), hybrid bonding is expected to reach only about 50% of HBM output; and by 2028 the entire hybrid bonder category is projected to approach USD 2 billion in market size. Set that last number beside BESI’s market capitalization of about EUR 24.38 billion: one company is valued at roughly ten times the annual market size of the entire equipment category three years out. The market is plainly not pricing 2028 hybrid bonding revenue; it is discounting a further-out endgame in which hybrid bonding becomes the standard process for all advanced packaging — which is the definition of option premium.
The competition deserves a full accounting too. ASMPT has broken into SK hynix’s bonding equipment supply chain and plans to ship next-generation tools to HBM customers; Hanwha Semitech’s second-generation hybrid bonder is slated for volume production in early 2026; Hanmi Semiconductor’s own hybrid bonding product is planned for launch by the end of 2027; LG Electronics is entering from outside with a 2028 production target (industry supply-chain analysis). On the same HBM back-end line, Hanmi sat on the incumbent generation’s throne and watched the HBM4 installed base get split down the middle by ASMPT (about 50 tools installed, roughly 25 each, per TrendForce and The Elec — see the Hanmi piece); BESI holds the leading position in the next generation with the earliest volume qualification. Structurally, BESI is far more diversified — 20 customers across four domains, with no Hanmi-style single-customer cliff. But the lesson of the Hanmi piece applies here too: bonding equipment share is generational, and thrones are most fragile exactly at generation changes and ramp points. BESI’s lead today is a lead in a market that has not yet scaled; by the time HBM volume truly arrives, everyone will be at the table for the carve-up.
AMAT’s 9% stake has carried a double identity since the day it was announced: a reinforcement of a five-year technology alliance, and a potential card in an acquisition game. The March 13, 2026 reports put the second identity on the table — both Lam Research and AMAT reported to be weighing a bid, with BESI retaining Morgan Stanley (Investing.com, IndexBox; all reported-rumor basis, no formal announcement). The market completed the first repricing the same day: up 10.2%.
The pricing difficulty with this option is that it is entangled with the first one. Within the roughly 63% gain from March 13 to June 12 sit the first-quarter order surge, the strong second-quarter guidance, and the fermenting bid expectation — there is no clean decomposition. The most direct proxy available is the target-price gap: EUR 317 stands about 25.5% above the 23-analyst median of EUR 252.55. The sell side’s earnings models do not produce the current price; the candidate explanations for the gap are either a bet on substantial upward earnings revisions, or a control premium. The proxy deserves a haircut: the update dates of individual targets cannot be verified one by one, and some may lag June’s tape.
The option has another side. When AMAT announced its stake, it explicitly stated it would not seek a board seat and did not plan to buy more — the posture of a strategic partner, not an acquirer. And that 9% is a wall in front of any third-party bidder: an offeror would face a direct competitor holding the largest outside stake. Add the multi-jurisdiction antitrust scrutiny that equipment-sector consolidation typically attracts (a general observation, not case-specific information), and this option could easily go unexercised for a long time. If the rumor cools, the 25.5% target-price gap is exposed: before March 13, this stock traded around EUR 176 (backing out the rumor-day move, a rough calculation).
Beneath the two options lies the floor. BESI’s core business — die attach, flip chip, multi-module placement, serving mobile, automotive, and industrial end markets — has been flat for three years: fiscal 2023 revenue of EUR 578.9 million, fiscal 2024 at EUR 607.5 million (up 4.9%), fiscal 2025 at EUR 591.3 million (down 2.7%). Mobile, automotive, and industrial demand remains depressed, with company-level results held up by AI and photonics applications (company IR).
This depressed base deserves a fair word on quality, though. Gross margin has held above 63% for three consecutive years (64.9%, 65.2%, 63.3%); the 1.9-percentage-point decline in fiscal 2025 was attributed by the company mainly to the roughly 12% first-half depreciation of the dollar against the euro — a currency headwind, not price competition. Net income fell much more steeply — from EUR 182.0 million to EUR 131.6 million, down about 27.7%, against a revenue decline of only 2.7%; the gap traces to currency and tax items, with no complete public breakdown (filed under blind spots). The dividend policy spells out the cash-cow character: payout ratios of 95% in both fiscal 2024 and 2025, with dividends per share of EUR 2.18 and EUR 1.58 — whatever is earned gets paid out. In management’s hands, the mainstream business is a dividend machine.
The inflection signal arrived in the first quarter of 2026: revenue up 28.3% year over year, net income of EUR 51.6 million, up 63.8%, with net margin expanding from 21.9% a year earlier to 27.9%; the company cited higher shipments to high-end mobile; second-quarter guidance calls for another 30% to 40% sequential gain. At its 2025 investor day, the company cited a TechInsights forecast of 42% cumulative growth from 2026 through 2029, driven by AI, new product cycles, and a mainstream recovery.
The valuation arithmetic only makes sense read against this floor. The trailing price-to-earnings ratio of 148 to 161x has as its denominator trailing-twelve-month net income of about EUR 150 million, which already includes the first quarter of 2026 — the weakest year-ago quarter has been swapped out for a recovered one, so the “suppressed denominator” holds for the three fiscal-2025 quarters inside it: real, but less extreme than it looks; the forward multiple of about 68x already builds the V-shaped earnings recovery into the estimate — and it still stands at two and a half times the company’s own 10-year median of about 27x. Put differently: even if second-quarter guidance is delivered in full and the earnings repair arrives on schedule, the current price still embeds an expectation of “recover first, then double again.”
Now the subtraction.
Layer one: mainstream cash flow. Three years of flat revenue, a 95% payout, gross margins above 63% — a high-quality machine that does not grow. Priced at the company’s own valuation habit of a roughly 27x 10-year median, this layer sits more than five times below the current 148 to 161x. The floor is real, but it does not explain the height of the building.
Layer two: hybrid bonding. On the evidence that belongs specifically to this layer — unit orders more than doubling sequentially in a single quarter, clearing the 2024 peak, customers up from 18 to 20 — the probability of this option paying off is rising, and that is real; the doubling of total orders and backlog is a company-wide print with the first layer’s recovery mixed in, and does not get booked here. But 2026 is the qualification year on the HBM side, volume production comes after 2027, and penetration reaches only about half by the HBM4E era. Over the next two years, hybrid bonding gives BESI orders and a narrative, not yet the main body of the income statement — in today’s price, this layer is mostly option premium, not delivered cash flow.
Layer three: the takeover. The 25.5% target-price gap is its most direct proxy; the 10.2% single-day move on March 13 is its cleanest standalone pricing. Its character is binary: an offer appears and the premium pays; the rumor cools and the premium goes to zero. But this premium differs in kind from the hybrid bonding one: the further the stock falls, the cheaper the target and the likelier a bid — and a 9% strategic holder plus a retained Morgan Stanley give this layer a measure of self-stabilization.
One qualifier has to go on record: BESI does not disclose hybrid bonding’s share of revenue (the company describes it only as a significant driver), so the three layers above are a qualitative decomposition, not a quantitative attribution — no one can compute from public data exactly how much of EUR 317 each layer accounts for. What can be said is structural: of the three layers, only the first is delivered cash flow, and the distance between its corresponding valuation habit and the current price has to be filled by the two options. Options can be very valuable — so far in 2026 they certainly have been — but the option-premium part of the price splits in two: the hybrid bonding share has, by its nature, no floor; the takeover share carries the self-stabilization described above — a soft cushion, provided the rumor is not formally denied.
This piece has leaned skeptical, so by our convention the other side gets its full strength.
First, orders are not narrative; they are purchase contracts. Total orders doubled year over year for two consecutive quarters, backlog more than doubled within a quarter, and unit orders cleared the 2024 peak — the first two are company-wide figures with mainstream recovery mixed in — the cut made above; what belongs to hybrid bonding is the third. The “option premium” framing understates the certainty: the inflection is already in the order book, not in a slide deck.
Second, the earnings repair has already started. First-quarter revenue up 28.3% and net income up 63.8%, with guidance for another 30% to 40% sequential gain — a forward multiple of 68x gets digested quickly by upward revisions over the coming quarters. Cyclical equipment stocks carry their highest price-to-earnings ratios at the bottom of the cycle and their lowest at the earnings peak; a 148x trailing multiple says the denominator is at its trough, and using it to argue bubble is using the wrong ruler.
Third, the moat is engineering, not first-mover luck. Alignment tolerance within 1 micron, a fully integrated system co-developed with AMAT over five years, volume qualification on TSMC’s SoIC — this is not a lead held together by being early. The script from the Hanmi piece — a TC throne breached within a year — does not transfer cleanly: the breacher there, ASMPT, was already a mature equipment giant, while hybrid bonding’s chasers have yet to show a single public volume qualification among them.
Fourth, the customer structure has no cliff. Memory is pure increment — evaluation tools have shipped to a second memory customer, and three are in evaluation or deployment. Against Hanmi’s structure of more than 80% of revenue tied to one customer, BESI is among the most customer-diversified assets on this line.
Fifth, AMAT’s 9% is a value signal verified with real money. The strategic buyer that best understands front-end equipment took a stake in April 2025; in March 2026, two potential buyers were reported to be circling the same asset — the world’s only listed pure play on hybrid bonding. Scarcity itself earns option premium. And the self-stabilization described a section earlier reads, on the bull side, as a positive: the lower the stock falls, the cheaper the target and the higher the odds of a bid.
Sixth, the 27x ten-year median may itself be a broken anchor. It is the valuation habit of the mainstream assembly-equipment era; if the structural switch holds — hybrid bonding growing from option layer into the main body of the income statement — the historical median measures the company that was, not the one taking shape. Arguing bubble from the old anchor quietly assumes the switch never happens — that is the re-rating case.
The common weakness of the bull case is the same as in the Hanmi piece: every plank awaits verification. Orders must become revenue, evaluation tools must become production orders, rumors must become offers — bulls and bears are working off the same calendar.
The brief behind this piece leaves nine data gaps, all laid out here. A clean year-to-date 2026 return could not be confirmed (source conventions conflict), so this piece uses the one-year gain of 167.6% instead. BESI never discloses its hybrid bonding customer list; whether the memory side includes SK hynix or Samsung is inference without a citable source, so this piece names only TSMC, AMD, Broadcom, and Apple, which have reporting behind them. Whether AMAT’s 9% has changed since April 2025 has no disclosure on record. The full breakdown of the gap between fiscal 2025’s 27.7% net income decline and its 2.7% revenue decline (currency versus tax) is not available in public filings reviewed. EV/EBITDA, operating cash flow, and capital expenditure could not be obtained as structured figures — so the cash-flow support behind the 95% payout ratio cannot be independently verified. Hybrid bonding’s revenue share is undisclosed, which is why the three-layer decomposition here can only be qualitative. The third quarter of 2025’s standalone order figure is missing. Developments in the takeover story between March and June are unknown. CEO Richard Blickman has served for more than 30 years with no public succession arrangement.
Two source conflicts also need flagging. First, on TrendForce’s account a hybrid bonder sells for about 4 billion won per unit (about USD 2.9 million) and above a TC bonder — but the ASMPT TC bonder order verified in the Hanmi piece (7 units for about 30 billion won) works out to roughly 4.3 billion won per unit. The two accounts do not reconcile, so this piece does not adopt the claim that hybrid bonders carry a clearly higher unit price than TC bonders. Second, the 52-week range supplied by data sources (EUR 103.25 to 291.10) contradicts the current price of EUR 317 and appears stale; it is not cited.
This report is an independent KSINQ market observation for informational purposes only. It is not investment advice. All acquisition interest described here is media-reported rumor with no formal offer announced as of publication; analyst targets are relayed from third-party data aggregations. Data snapshot: June 13, 2026; price data as of the June 12, 2026 close.