Besi’s beat is less about the quarter than the order-to-revenue lag the market is underpricing

The central read from this print is that Besi is transitioning from a cyclical assembly-equipment recovery story into a more concentrated advanced-packaging conversion story, and the street may be mispricing the timing rather than the demand. What was priced in was a modest beat against revenue of €191.1 million and EPS of €0.59, consistent with a recovery still constrained by mainstream mobile, automotive, and industrial weakness. What actually surprised was that street-basis revenue came in at €195.7 million, a +2.4% surprise, while EPS came in at €0.63, a +7.1% surprise, despite the company’s own historical margin series still showing the scars of Q3 FY2025, when gross margin fell to 58.0% and EPS fell to €0.31. The variant perception is that investors should not over-anchor on the apparent moderation embedded in the next-quarter revenue cadence, because the Q4 order surge was back-end loaded and tied to long-throughput tools, not ordinary short-cycle die attach demand. Richard Blickman’s most useful sentence on the call was not promotional, it was operational: “The orders were -- or let's say, the order placements were very much to the end of the quarter and the manufacturing throughput time for many of these orders, so take the high-end Flip Chip machines, the CHAMEOs, but also the multi-module attach, which is very much also for photonics, they take 12 to 16 weeks.” That wording matters because it explains why Q4 demand can be real while Q1 revenue recognition is not yet fully reflecting it.

That timing point matters more because the financial trajectory had already turned before the order book fully converted. The quarterly history shows revenue bottoming at €132.4 million in Q3 FY2025, then rebounding to €164.8 million in Q4 FY2025, with revenue QoQ moving from -10.6% in Q3 FY2025 to +24.4% in Q4 FY2025. Gross margin followed the same repair path, rising from 58.0% in Q3 FY2025 to 63.9% in Q4 FY2025, which is the margin evidence behind the EPS recovery from €0.31 to €0.54 in the company’s quarterly history. The street-comparison basis was even better on the reported event, with actual revenue of €195.7 million versus estimate €191.1 million and actual EPS of €0.63 versus estimate €0.59. The separation is important: use the street basis for the +2.4% revenue surprise and +7.1% EPS surprise, and use the company’s own accounts when discussing management’s operational bridge. On that basis, the company’s disclosed Q1 FY2026 history shows revenue of €187.9 million, gross margin of 60.3%, revenue QoQ of +14.0%, revenue YoY of +30.4%, and diluted EPS of €0.66, which supports the thesis that demand is accelerating faster than margins have yet normalized.

The revenue chart should force the debate away from whether Besi printed one clean quarter and toward whether the gross-margin dip at €187.9 million of Q1 FY2026 revenue is a mix and timing issue rather than a structural reset. The company’s own Q4 FY2025 gross margin of 63.9% was near the Q4 FY2024 level of 64.0% and above Q1 FY2025 at 63.6% and Q2 FY2025 at 63.3%, while Q1 FY2026 gross margin of 60.3% sits below that repaired run rate even as revenue YoY was +30.4%. That is the conflicting evidence to respect: revenue and EPS argue demand strength, while Q1 FY2026 gross margin argues that mix, ramp costs, or shipment composition can still dilute conversion. But the company guided for gross margins ranging between 63% and 65% for the first quarter ’26, and the Q4 call language tied that to “improved revenue and a more favorable advanced packaging product mix.” If gross margin prints inside 63% to 65% on the next reported quarter while revenue follows the order book, the current concern about margin volatility should fade; if gross margin remains closer to 60.3%, the advanced-packaging thesis will need a lower near-term earnings multiple.

The order data is the reason to lean toward conversion rather than fade the move. Management said Q4 orders were €250.4 million and increased by 43.3% versus the third quarter of ’25, driven by Asian subcontractors for 2.5D data center applications, renewed photonics capacity purchases, and hybrid bonding orders. For the full year, orders were €685 million and increased by 16.8% versus 2024, but the more important composition signal is that growth accelerated in the second half, with orders increasing 63.6% versus the first half of ’25. That is not just a cyclical upturn in unit assembly tools. Orders for AI applications represented approximately 50% of total orders in ’25, and computing end-user revenue grew from approximately 40% of revenue in 2024 to 50% in 2025. The market may be treating Besi as still hostage to broad assembly weakness because annual revenue of €591.3 million decreased by 2.7% versus 2024, but the order mix says the next revenue cycle is being pulled by a different end-market set than the one that depressed 2025 shipments.

That mix shift also explains why the next-quarter guide can look underwhelming next to Q4 orders without actually undermining the thesis. The company’s own call basis had Q4 revenue at €166.4 million and orders at €250.4 million, while the first-quarter ’26 outlook called for revenue to increase between 5% and 15% versus the fourth quarter of last year. An analyst challenged that apparent gap directly by asking why, on €250 million in Q4 orders, the company was guiding to only “€185 million or so at the midpoint.” The answer was not demand caution, but late-quarter order placement and 12 to 16 weeks of manufacturing throughput for high-end Flip Chip, CHAMEO, and multi-module attach systems. That matters for portfolio positioning because it shifts the burden of proof from “why was the guide not higher?” to “does the backlog convert across the following 12 to 16 weeks?” The market often discounts order spikes when revenue guidance does not immediately follow, but here the quoted throughput window provides a testable reason why shipment timing lags order intake.

The second-order implication for the customer chain is that AI packaging capacity is moving from planning language into tool-line deployment, even though the data pack does not identify named Besi customers. Asian subcontractors are explicitly called out as the source of a broad-based increase in demand for 2.5D data center applications, and a leading logic customer installed 6 integrated hybrid bonding production lines incorporating 30 Besi-hybrid bonders in collaboration with Applied Materials. Those are the magnitudes that matter: 6 lines and 30 bonders indicate that hybrid bonding is no longer a single-tool qualification story at that customer, while Q4 orders of €250.4 million and full-year orders of €685 million show the demand is not isolated to one installation. Applied Materials is the named partner with the cleanest read-through: the collaboration embeds Applied Materials in integrated hybrid bonding production lines at a leading logic customer, so Besi’s 30-bonder deployment points to process-flow integration rather than standalone bonder shipments. For Asian subcontractors, the read-through is capacity pull for 2.5D data center applications, not broad OSAT recovery, because management tied the Q4 order increase specifically to Asian subcontractors and 2.5D data center demand.

The supplier read-through is limited by the data pack, because no suppliers to Besi are listed, and that absence is itself relevant for how far one should push second-order conclusions. The print supports read-throughs to Asian subcontractors, a leading logic customer, and Applied Materials because they are explicitly named or described in the call excerpts, with magnitudes of €250.4 million in Q4 orders, 6 installed production lines, and 30 hybrid bonders. It does not support naming component suppliers or extrapolating shortages in precision stages, optics, or substrates, because the supply-chain section lists no suppliers to Besi. The better inference is therefore demand-side capacity allocation, not supply tightness. Management also said automotive was between 15% and 20% of revenue, which helps frame what is not driving the beat: automotive is meaningful, but the order acceleration was attributed to AI 2.5D, photonics, and hybrid bonding rather than automotive recovery.

The peer context reinforces that Besi is not the only test-and-assembly name seeing growth, but its margin and mix story are different from a simple subsector beta trade. In the latest peer table, ATEYY reported revenue YoY of +43.8% with gross margin of 67.4%, DSCSY reported revenue YoY of +12.3% with gross margin of 70.8%, and 6871.T reported revenue YoY of +48.3% with gross margin of 47.3%. Against that backdrop, Besi’s company-history Q1 FY2026 revenue YoY of +30.4% and gross margin of 60.3% place it below the highest-margin Japanese peers but above lower-margin assembly-equipment peers such as 7729.T at 42.4%, 6315.T at 36.2%, 6941.T at 38.0%, and 6140.T at 28.3%. The actionable comparison is not that Besi has the best reported margin in the table; it does not. The point is that Besi has a high-margin equipment model with an order mix tied approximately 50% to AI applications in ’25, and that combination should command more attention than the annual revenue decline of 2.7% versus 2024 would imply in isolation.

The call tone supports the same interpretation, but it is useful because it improved in delivery while confidence fell, which is a better setup than one-sided promotional language. The tone history shows Q1 FY2026 sentiment at 0.33 versus 0.24 in Q4 FY2025, guidance_tone at 0.72 versus 0.27, prepared_sentiment at 0.92 versus 0.70, and uncertainty down to 33.2 from 46.5. Q&A also improved, with qa_sentiment at 0.20 versus 0.13 and qa_evasiveness at -32.0 versus -25.7. The caveat is that tone_confidence fell to 0.30 from 0.35, so the transcript became more positive and less uncertain, but not more statistically confident. That is a subtle but important balance: management sounded clearer on demand and guidance, while the model’s lower tone_confidence argues against treating the transcript as a standalone buy signal without order conversion evidence.

That tone shift is most investable when tied to the hard order and margin numbers rather than the adjectives on the call. Guidance_tone rose by +0.45 call-over-call, uncertainty fell by -13.3, and qa_evasiveness fell by -6.3, which fits a management team that has moved from explaining weak mainstream assembly to discussing longer-lead advanced-packaging shipments. Yet ai_optimism only rose by +0.05, from 0.28 to 0.34, even though orders for AI applications represented approximately 50% of total orders in ’25. That gap is the variant perception in language form: the transcript did not simply hype AI; it gave operational bottlenecks, throughput timing, and installed-line detail. Blickman’s statement that “Customer road maps also point to expanded adoption of wafer-level assembly over the next 2 years related to hybrid bonding and TC NXT adoption in HBM 4, 4E, co-package optics, ASICs and new high-performance computing and mobile introductions” matters because it commits to specific packaging vectors over a 2 years horizon, not just a generic AI capex cycle.

The profitability and capital-return backdrop make the timing risk tolerable, but not irrelevant. For 2025, the company reported gross, operating, and net margins of 63.3%, 29.3%, and 22.3%, respectively, while liquidity at year-end included cash and deposits of €543 million and net cash of €36 million. Net cash increased by €43.8 million versus September 30, ’25, and cash and deposits increased by €24.4 million versus September 30, ’25. The proposed cash dividend is €1.58 per share, representing a 95% payout ratio, which means management is not conserving cash as if the order book were speculative. Still, the payout also reduces the margin for operational disappointment if advanced-packaging capacity requires higher development spending. Management guided operating expenses to increase by 10% to 15% as development spending rises, so investors should not assume every euro of revenue upside drops through at Q4 FY2025 incremental margins. The thesis is order conversion plus mix repair, not cost starvation.

What to watch next is therefore concrete. First, revenue must track the company’s first-quarter ’26 guide for a 5% to 15% increase versus the fourth quarter of last year, and the street should judge any shortfall against the 12 to 16 weeks throughput explanation rather than against the Q4 order headline alone. Second, gross margin needs to land in the guided 63% to 65% range; a result nearer the Q1 FY2026 historical 60.3% would break the argument that mix is normalizing. Third, order quality matters more than order size: AI applications were approximately 50% of total orders in ’25, computing end-user revenue moved from approximately 40% of revenue in 2024 to 50% in 2025, and second-half orders increased 63.6% versus the first half of ’25, so the next call needs evidence that those mix markers are holding rather than reverting to weaker mobile, automotive, and industrial demand. Finally, the April AGM vote on the €1.58 per share cash dividend with a 95% payout ratio is a balance-sheet confidence check; if management maintains that payout while reiterating 63% to 65% gross margin and confirming backlog conversion from the late Q4 orders, the market’s concern about a soft guide should give way to the more important conclusion that Besi’s advanced-packaging cycle is arriving with a measurable lag, not fading.