Micro-Mechanics (Holdings) Ltd. (SGX: 5DD) - Deep Dive Research Report

Prepared 2026-06-18. Fiscal year ends 30 June. Most recent reported period: 3QFY2026, quarter ended 31 March 2026, released 28 April 2026.

Section 1: What the Company Does

Micro-Mechanics makes the tiny, high-wear consumable tools and precision parts that semiconductor factories use to handle and assemble chips. When a silicon wafer is cut into individual dies and each die is picked up, placed onto a lead frame or substrate, and wired up, the machines doing that work grip and manipulate parts smaller than a grain of rice, thousands of times an hour. The little rubber-and-carbide tips that touch the chip, the collets that pick the die, the capillaries that thread the bonding wire, and the cavity moulds that encapsulate the finished package all wear out and must be replaced constantly. Micro-Mechanics designs and manufactures those parts.

The founder, Chris Borch, came to Singapore in 1981 to work for an American maker of semiconductor assembly equipment (he had previously been at Kulicke & Soffa, one of the industry's pioneer equipment builders). He noticed that the chip-assembly industry building up across Southeast Asia had no good local supplier for the precision consumable tools its machines depended on, and in 1983 he started Micro-Mechanics with S$1,200 and one employee in the back of a hat-sewing factory. Over four decades it grew into a group with five factories - Singapore, Malaysia, China, the Philippines and Silicon Valley in the United States - serving more than 600 customers worldwide. The company listed on SGX in 2003; total shareholder return since listing exceeds 3,000%.

Borch's own analogy captures the business better than any segment chart:

"We supply the critical performing parts for systems - like producing sewing needles for sewing machines or inkjet cartridges for printers."

That is the heart of the value proposition. The machine (the sewing machine, the printer, the bonder) is a one-time capital purchase. The consumable part (the needle, the cartridge, the die collet) is bought again and again for as long as the machine runs. Micro-Mechanics sells the razor blades, not the razor. A die collet might cost a few dollars, but a single chip-assembly line burns through thousands of them, and a worn or out-of-spec tool damages expensive chips, so customers care far more about reliability and precision than about price. That is what makes the product hard: these tools are machined to micron tolerances, often from tungsten carbide, hardened steel, or proprietary high-temperature elastomers, and they have to grip a fragile silicon die without chipping it, while surviving thousands of thermal and mechanical cycles. Getting the geometry, the material, and the surface finish right took the company decades to learn, and each customer's chip package requires a custom tool design.

A concrete walk-through. A chip-assembly (OSAT) plant in Malaysia runs die-bonders that take dies off a diced wafer and place them onto lead frames. The pick-up head needs a rubber tip or a tungsten-carbide collet shaped precisely to the die's dimensions. Micro-Mechanics designs that collet to the customer's die geometry, machines it, and ships it. The customer qualifies it on the line, and once it works, reorders the same part on a recurring basis as tools wear out. Multiply that across wire-bonding capillaries, die-ejector needles, encapsulation mould components, and test-socket parts, across hundreds of customers and thousands of distinct package types, and you have the business.

Section 2: Business Segments

Micro-Mechanics runs two reporting segments that share precision-machining DNA but serve different points in the chip-making flow.

Consumable Tools (the core, roughly 82% of revenue)

This is the original and dominant business. It designs and manufactures the miniature consumable tools used in the back-end assembly and test of semiconductors: die-attach pick-up tools (rubber tips, high-temperature plastic tools, tungsten-carbide die collets, vacuum wands, sensor assemblies), die-ejection tools, dispensing tools, wire-bonding products, and encapsulation tooling. In 3QFY2026 the consumable tools segment grew 20.9% year-on-year to S$14.4 million and made up the large majority of group revenue (82.2% of revenue in 1QFY2026).

The core capability is process knowledge accumulated across forty years: knowing exactly what tip geometry and material will pick a particular die without cracking it, and being able to manufacture that tool repeatably to micron tolerances at a price low enough that it is bought as a consumable. The segment exists at scale because the assembly tools are genuinely consumable - they wear out fast and must be reordered - which gives the business a recurring, high-frequency order pattern more like an industrial razor-blade business than a capital-equipment one. This is the group's margin engine and cash cow: it carries the high gross margins (group gross margin reached roughly 50-52% in FY2026) and funds the dividend.

Wafer Fabrication Equipment (WFE) Parts (roughly 18% of revenue)

This segment, anchored in the US subsidiary Micro-Mechanics Inc. (MMUS) in Silicon Valley, makes precision parts and assemblies for the front-end - the machines that build chips on the wafer before they are ever cut and packaged. Where consumable tools are sold to chip assemblers, WFE parts are sold to the makers of wafer-fabrication equipment (the equipment OEMs) and serve customers based mainly in the USA, Singapore and Malaysia. This is contract precision machining and assembly of higher-complexity, higher-mix parts. In 3QFY2026 WFE sales were S$4.2 million, up 2.6% year-on-year; the segment is lumpier than consumables and is exposed to the fab-equipment capex cycle (it fell 15.3% in 1QFY2026 on material delays and shortages, then recovered).

The core capability here is complex multi-axis machining and assembly to semiconductor-grade cleanliness and tolerance, serving an entirely different customer base (equipment OEMs rather than chip assemblers) with different economics (project- and order-driven, lower margin, more cyclical). It exists as a separate entity largely because of where the customers are - WFE equipment OEMs cluster in Silicon Valley, so the company put a plant there. Management has framed WFE as the strategic growth bet and the bridge to becoming what it calls a "Next-Generation Supplier serving at the intersections of both the consumable tools and WFE segments." MMUS was a loss-maker for years; a restructuring that cut roughly S$1.3 million of annual operating costs turned it to its first full year of profitability in FY2025, which management treats as validation of the turnaround.

Section 3: Products and Business Detail

The product catalogue follows the chip-assembly flow. In die-attach pick-up, the company makes rubber tips, high-temperature plastic tools, tungsten-carbide die collets, vacuum wand tools and sensor assemblies - the parts that physically lift a die off the wafer and place it. In die ejection, it makes the ejector needles and tools that push the die up from below during pick-up. In dispensing, it makes the tools that lay down the die-attach adhesive. In wire bonding, it makes capillaries and related bonding tools that thread fine gold or copper wire from the chip to its package leads. In encapsulation, it makes the precision mould components used to seal the finished package in plastic. Alongside these catalogue consumables, the WFE arm does contract manufacturing of precision parts and tools for capital equipment makers.

What makes the products hard is the combination of materials science and micro-machining. Collets and capillaries are made from tungsten carbide, hardened tool steels, ceramics and proprietary high-temperature elastomers; each must hold its geometry under repeated thermal and mechanical stress while handling a die without chipping it. R&D in FY2026 has focused on advanced-packaging elastomers and new machining technologies, reflecting the industry's shift toward advanced packaging (chiplets, stacked die, heterogeneous integration) where tool geometries are evolving. Each tool is effectively bespoke to a customer's die and package, so the business is high-mix, low-unit-cost, and design-intensive.

Manufacturing sits across five factories: Singapore (headquarters and core consumables), Malaysia, China (Suzhou), the Philippines, and the United States (MMUS, Silicon Valley). This geographic spread is itself a selling point in the current environment, because customers increasingly want supply close to their own assembly sites and want to de-risk single-country sourcing; management has explicitly tied capex to "supply chain localisation" and "localised capabilities across the Group's five factories." The company runs an internal operational-excellence programme it calls "8S," scored on a five-star scale, and reported that all factories reached 4.5-star or better in FY2026 - the "Five-Star Factory" vision the new CEO is driving. Recent operational milestones include the MMUS turnaround to profitability (FY2025), ongoing cloud migration and cybersecurity hardening, and capex earmarked for advanced-packaging and wafer-fab-equipment manufacturing capability (around S$2.0-2.3 million planned for 2HFY2026, part of roughly S$4.0 million of growth-and-replacement capex flagged for FY2026).

Section 4: Customers

The customers are the world's semiconductor manufacturers and their equipment suppliers. For consumable tools, the buyers are chip assembly-and-test operations: outsourced assembly and test houses (OSATs) and the assembly lines of integrated device manufacturers (IDMs), spread across Asia and beyond. The company serves more than 600 customers, and in 1HFY2026 its revenue split geographically with China the largest at about 36%, the USA about 20%, and Malaysia about 18%, with Singapore and others making up the rest. For WFE parts, the buyers are wafer-fabrication equipment OEMs, concentrated in the USA, Singapore and Malaysia.

The buying decision for consumable tools is made by process and manufacturing engineers on the customer's assembly floor, not by a procurement department chasing the lowest price. Their criteria are tool life, dimensional precision, and consistency, because a tool that drifts out of spec or chips a die destroys product worth far more than the tool itself. The sales relationship is technical and iterative: the company co-designs a tool to a customer's specific die and package, the customer qualifies it on the line, and once qualified it becomes the reorder default.

That qualification step is the switching cost. Once a tool is designed-in and proven on a production line, swapping to a competitor means re-qualifying, re-testing, and risking yield on a running line - friction that engineers avoid unless there is a real problem. The tools are cheap relative to the chips they handle and the machines they run on, so there is little incentive to chase a marginally cheaper supplier. That is what gives the consumables business its stickiness and recurring order pattern.

Concentration is low and is a feature, not a bug: with 600-plus customers and a high-frequency consumable reorder model, no single customer dominates, which makes revenue more predictable than a project-based equipment supplier. The trade-off is that the business is still tied to overall industry utilisation - when chip assembly volumes fall in a downturn, tool consumption falls with them. Contract structure is mostly recurring purchase-order business rather than long-term take-or-pay contracts; predictability comes from the installed base of qualified tools and the consumable reorder cadence rather than from contractual lock-in. WFE parts, by contrast, are more order- and project-driven and therefore lumpier.

Section 5: Competitive Landscape

The consumable-tools market for semiconductor assembly is a specialised niche dominated by a handful of focused precision-tool makers rather than by large diversified industrials. Micro-Mechanics competes primarily against Small Precision Tools (SPT Roth) of Switzerland, a privately held pioneer in bonding capillaries and die collets with its own ceramic injection-moulding and material technology; PECO of South Korea, which supplies capillaries, pick-up tools, clamps and epoxy tools; and Micro Point Pro (MPP) of Israel, which makes die collets and pick-up tools. In bonding capillaries specifically, equipment maker Kulicke & Soffa also sells expendable tools and overlaps with parts of the range. In the WFE precision-parts business, MMUS competes against a fragmented field of precision machine shops plus larger listed contract manufacturers of semiconductor-equipment subsystems such as Ultra Clean Holdings and Ichor Holdings.

Micro-Mechanics wins on a combination of design responsiveness, the breadth of its qualified installed base, geographic spread across five countries (which matters more as customers localise supply), and consistently high gross margins that signal pricing power in its niche. Where it is exposed: against SPT in particular it faces a credible, technically deep, privately funded rival that can match it on materials science; in WFE it is a relatively small player against larger, better-capitalised subsystem suppliers and a long tail of regional machine shops, which is part of why that segment carries lower margins and more cyclicality.

Barriers to entry in consumables are real but not insurmountable: the hard part is not a single patent but accumulated process knowledge, the library of qualified tool designs across thousands of package types, and the customer qualification lock-in. A new entrant would have to reproduce decades of design know-how and then get each tool re-qualified line by line - slow and unrewarding for a low-ticket consumable. That keeps the field small and stable rather than producing a single dominant monopolist; share is distributed among the focused specialists named above.

Market-cap figures are rough peer-size references as of June 2026 and move with the market.

Section 6: Industry

Demand for Micro-Mechanics' products is driven by the volume of chips being assembled, tested and fabricated worldwide - not by the price of chips, but by the unit throughput of the back-end and front-end. When assembly lines run hot, they consume more tools; when fabs add capacity, they buy more precision parts. The proximate drivers right now are the AI and data-centre build-out and the resulting strength in Logic and Memory, which management has repeatedly cited as the source of demand momentum across all regions. The broader semiconductor market is forecast to reach roughly US$975 billion in 2026 (management's cited figure), and the wafer-fab equipment market that the WFE segment serves is itself a large and growing pool, projected by industry researchers (Yole) toward US$165-184 billion later this decade.

Micro-Mechanics sits in the picks-and-shovels layer of this supply chain. It does not make chips and does not make the big machines; it supplies the consumable tools and precision parts that both depend on. That position has two consequences. First, it is geographically tied to where assembly and fabrication actually happen - hence five factories spanning Singapore, Malaysia, China, the Philippines and Silicon Valley, and a deliberate "localisation" capex strategy as customers regionalise their own supply chains amid US-China trade tension. Second, it is cyclical: the semiconductor industry runs in capex and inventory cycles, and the company felt the 2023 downturn clearly (the dividend was cut from 9.0 to 6.0 cents) before recovering strongly through FY2025 and FY2026. The consumables business dampens that cyclicality relative to pure equipment makers, because tools are reordered continuously, but it does not eliminate it.

Regulation in the conventional sense (product approvals) is light, but the industry is increasingly shaped by export controls and trade policy, particularly US-China semiconductor restrictions, which both create localisation demand (a tailwind for a multi-country supplier) and risk fragmenting the China end-market (where the company earns its single largest geographic slice of revenue). The current industry backdrop is a tailwind: AI-driven demand, a forecast upturn across Logic and Memory, and supply-chain localisation all favour a diversified consumable-tool supplier. The headwind is the ever-present risk of the next cyclical inventory correction and the geopolitical overhang on China exposure.

Section 7: Growth Triggers

All of the following are drawn from the company's quarterly results commentary across the six most recent reporting periods.

• Broad-based demand uptick across all regions and product segments, led by Logic and Memory on AI demand (3QFY2026, 28 Apr 2026; repeated from 2QFY2026, 29 Jan 2026). Management has carried this theme across multiple quarters.

  "Management expects to benefit from an uptick in demand across all regions and product segments, particularly in Logic and Memory." (3QFY2026)

• Advanced-packaging capability investment. Capex of roughly S$2.0 million in 2HFY2026 directed specifically at advanced packaging and wafer-fabrication-equipment manufacturing, plus R&D into advanced-packaging elastomers (3QFY2026, 28 Apr 2026; 2QFY2026, 29 Jan 2026 flagged S$2.3 million).

• Supply-chain localisation across the five-factory network. Around S$4.0 million of growth-and-replacement capex flagged for FY2026 to enhance capacity and "localised capabilities across the Group's five factories, in line with the industry trend towards supply chain localisation" (1QFY2026, Oct 2025).

• MMUS / WFE segment recovery and scaling after its turnaround. Having reached first full-year profitability in FY2025, the WFE segment is positioned as the growth bet; WFE sales swung from a 95.7% year-on-year jump (3QFY2025, 29 Apr 2025) through a material-shortage-driven dip (1QFY2026) back to growth (3QFY2026).

• Five-Star Factory / 8S operational-excellence programme driving margin and productivity, with all factories reaching 4.5-star or better (3QFY2026, 28 Apr 2026), under the new CEO's continuous-improvement agenda.

• Industry growth toward roughly US$975 billion in 2026 cited as the demand backdrop, with data-centre and AI computing platforms as named drivers (3QFY2026, 28 Apr 2026; the figure was revised up from earlier US$728-800 billion estimates given at FY2025, Aug 2025).

Section 8: Key Risks

Semiconductor cyclicality feeding straight through to tool consumption. Micro-Mechanics' revenue tracks the volume of chips being assembled and the fab-equipment capex cycle. When the industry enters an inventory correction, assembly lines slow and tool reorders drop, and the WFE segment, tied to equipment capex, falls faster. This is a high-probability, moderate-to-significant drag risk: it is not a question of if but when the next downturn comes. The 2023 cycle already demonstrated it - the company cut its dividend from 9.0 to 6.0 cents - and even within FY2026 the WFE segment fell 15.3% in a single quarter on material delays.

China concentration colliding with trade policy. China is the company's single largest geographic market at roughly 36% of revenue. Tightening US-China semiconductor export controls could disrupt the China assembly end-market, redirect customer supply chains, or expose the company to retaliatory measures. The same trade tension that creates localisation demand (a tailwind) also threatens the largest revenue slice. This is a moderate-probability, potentially significant risk that is largely outside management's control.

WFE segment execution and lumpiness. MMUS only just reached its first full year of profitability in FY2025 after years of losses. The segment is lower-margin, more cyclical, and dependent on a smaller set of equipment-OEM customers, and it has already shown vulnerability to material shortages. If the WFE recovery stalls or reverses, the group loses its headline growth narrative and could see the segment slip back toward break-even.

Founder-to-second-generation leadership transition. Chris Borch, who built the company over 40 years, stepped down as CEO on 30 June 2025, handing operations to his son Kyle Borch (previously a manufacturing engineer at the company, with prior experience at Apple, Agilent and NASA JPL) while remaining Executive Chairman. Founder-led companies carry the risk that institutional knowledge and customer relationships are concentrated in the founder; a family succession mitigates continuity risk but introduces the question of whether the next generation can sustain the operational discipline and margins the founder built. This is a low-probability but structurally important risk.

Margin dependence on the consumables mix. The group's attractive gross margins (around 50-52% in FY2026) rest heavily on the high-margin consumables segment. A shift in mix toward lower-margin WFE work, or pricing pressure in consumables from rivals like SPT or PECO, would compress group margins. Management's own framing of becoming a supplier "at the intersection of both segments" implies a deliberate push into the lower-margin WFE business, which is a growth lever and a margin risk at the same time.

Section 9: Walk the Talk

The six reporting periods used for this assessment are: 2QFY2025 / 1HFY2025 (quarter ended Dec 2024, reported 26 Jan 2025); 3QFY2025 (Mar 2025, reported 29 Apr 2025); FY2025 / 4QFY2025 (year ended Jun 2025, reported 28 Aug 2025); 1QFY2026 (Sep 2025, reported Oct/Nov 2025); 2QFY2026 / 1HFY2026 (Dec 2025, reported 29 Jan 2026); and 3QFY2026 (Mar 2026, reported 28 Apr 2026). The most recent is within ~50 days of today's date.

Starting with the 2QFY2025 results in January 2025, management framed the story as a recovery off the 2023-24 downturn, posting 108.9% year-on-year net profit growth and pointing to the MMUS/WFE turnaround as the next leg. The promise embedded here was twofold: that the cost restructuring at the US subsidiary would carry it to profitability, and that semiconductor demand was inflecting upward.

By 3QFY2025 (April 2025), both were tracking. WFE sales jumped 95.7% year-on-year to S$4.0 million, contributing over a quarter of revenue, and net profit rose 72.6%. This was also the quarter the leadership transition was announced - Kyle Borch to take over as CEO on 1 July 2025 with Chris Borch becoming Executive Chairman - a succession handled in advance and on a clear date rather than abruptly.

The FY2025 results in August 2025 delivered the headline promise: MMUS reached its first full year of profitability, with the restructuring confirmed to have removed roughly S$1.3 million of annual operating costs. Group revenue grew 12.6% to S$65.2 million and EBITDA margin expanded from 31.8% to 34.9%. The dividend was held at 6.0 cents (67.3% payout), consistent with the stated policy of paying out at least 40% of earnings. Management delivered on the cost-out and turnaround commitments it had been making for several quarters.

"MMUS achieved its first full year of profitability, validating the effectiveness of prior restructuring efforts." (FY2025, Aug 2025)

The FY2026 quarters then tested consistency. In 1QFY2026 (October 2025) the WFE segment slipped 15.3% on material delays - an honest disclosure of a near-term setback - even as consumables (82% of revenue) kept growing and the company flagged S$4.0 million of localisation capex. By 2QFY2026 (January 2026) the group was back to 25.2% net-profit growth with gross margin up to roughly 51% from 47% a year earlier, and by 3QFY2026 (April 2026) revenue reached S$18.6 million with gross margin around 51.6% and a 26.3% nine-month ROE. The margin-expansion trajectory management had been describing materialised quarter after quarter.

On capital returns, the record is steady rather than aggressive: the dividend was cut once, decisively, in the 2023 downturn (9.0 to 6.0 cents) and then held flat at 6.0 cents through FY2024 and FY2025, with the policy explicitly stated and adhered to.

The overall read: this is management that does broadly what it says. The MMUS turnaround was promised across several quarters and then delivered on schedule; margin expansion was guided and achieved; the dividend policy was stated and followed, including an honest cut when earnings fell rather than an unsustainable hold. The disclosures are candid about setbacks (the 1QFY2026 WFE dip was not buried), and the leadership succession was executed in an orderly, pre-announced way. The main caution is that the favourable AI-demand narrative coincides with an industry upcycle, so the next downturn will be the real test of whether the discipline holds.

Section 10: Shareholder Friendliness Index

Dividends. Over the last three completed fiscal years the dividend was 9.0 cents in FY2023, then cut by a third to 6.0 cents in FY2024, and held flat at 6.0 cents in FY2025 (paid as 3.0 cents interim plus 3.0 cents final each year). The FY2024 cut was a direct response to the 2023 semiconductor downturn and falling earnings rather than a change of policy; the company maintains a stated policy of distributing at least 40% of after-tax earnings, and the FY2025 payout ratio was 67.3% - well above the floor, signalling a genuine commitment to returning cash even while reinvesting. Into FY2026 the company has declared 3.0 cents interim each quarter, consistent with the 6.0-cents-plus annual cadence. Cumulative dividends since the 2003 listing reached 137.9 cents per share by 3QFY2026.

Buybacks and dilution. Share repurchases are minimal and opportunistic rather than programmatic. The company carries a standard share-buyback mandate renewed at its AGM, but used it lightly: it reported no buybacks in FY2025, and in 3QFY2026 it repurchased just 26,300 shares for about S$50,000 - a token amount against roughly 141 million shares outstanding. The share count has been essentially flat over the three-year period; there is no meaningful option-driven dilution, and no large buyback programme has been executed. (MoatMap's database shows no buybacks in the trailing ~90-day window to 18 June 2026; the FY2023-FY2025 picture above is sourced from the annual results and exchange filings, and confirms repurchases have been negligible throughout.) Capital return is therefore overwhelmingly via dividends, not buybacks.

Verdict: Returns Capital - a consistent, policy-driven dividend payer (40%+ payout, ~67% in FY2025) that cuts honestly in downturns and holds steady otherwise, with buybacks immaterial to the story.

Section 11: Insider Activities

Per the MoatMap cross-market disclosure database (market: SG), which is the canonical source for recent insider dealing on this venue because the SGXNet disclosure portal is gated and external search returns auth-blocked stubs, there were zero insider transactions recorded for 5DD.SI over the trailing twelve months (data current as of 2026-06-18). Accordingly there are no director, officer, or substantial-shareholder open-market buys or sells to report for the period.

The absence of recorded transactions is itself a mild data point rather than a signal: there was no insider selling into the FY2026 share-price strength, but equally no conviction open-market buying to flag. The founder family (Chris Borch as Executive Chairman, Kyle Borch as CEO) retains a substantial controlling stake built over four decades, and the leadership transition in mid-2025 was a role handover within the family rather than a stake disposal, so the family's alignment with outside shareholders remains intact. Net assessment: neutral - no insider buying or selling activity in the last twelve months per the available disclosure feed, with founder-family ownership unchanged through the CEO succession.

Section 12: Scenarios

Bull case. The AI and data-centre build-out keeps Logic and Memory assembly volumes running hot across all of Micro-Mechanics' regions, and the consumable-tools business compounds steadily as its qualified installed base widens with each new package type. The advanced-packaging push pays off: as chiplets, stacked die and heterogeneous integration become mainstream, the company's new elastomer and tooling designs get designed into the next generation of assembly lines, and because advanced packaging uses more and finer tools, consumption per chip rises. MMUS, now profitable, scales into a genuine second engine as equipment OEMs localise their supply chains and reward a multi-country supplier, and the WFE margin gap narrows. Kyle Borch's Five-Star Factory programme keeps lifting productivity, holding group gross margins above 50%. The five-factory footprint turns the US-China trade fracture into an advantage rather than a threat, because customers want exactly the kind of regionally diversified, qualified supplier the company already is. Dividends grow again off a higher earnings base, and the founder-to-son transition proves seamless.

Base case. The semiconductor upcycle continues at a more measured pace. Consumables, at over 80% of revenue, keep growing in line with industry assembly volumes and remain the dependable margin engine and dividend funder. WFE stays lumpy - good quarters and soft quarters driven by equipment-OEM order timing and the occasional material shortage - but holds its newly won profitability and contributes modest growth. Group gross margins hold in the low-50s, the 8S programme delivers incremental rather than transformational productivity gains, and capex stays disciplined at a few million dollars a year directed at advanced packaging and localisation. The dividend is maintained around the 6-cents-plus level with the 40%-plus payout policy intact, and the new CEO runs the business with the same operational discipline as the founder. Nothing breaks; nothing dramatically surprises. China remains the largest market but trade policy stays manageable.

Bear case. The AI-driven upcycle rolls over into the next inventory correction. Chip-assembly volumes fall, consumable-tool reorders drop with them, and the WFE segment - tied to equipment capex and only recently profitable - slips back toward break-even or losses as OEM orders dry up and material shortages bite. China, the largest single market at over a third of revenue, gets squeezed by an escalation in US export controls or retaliatory measures, fragmenting demand and forcing costly supply-chain reconfiguration. Group margins compress as volume falls against a fixed five-factory cost base and as any mix shift toward lower-margin WFE work dilutes the consumables-driven profitability. The dividend is cut again, as it was in 2023, testing income-oriented shareholders' patience. In the background, the founder transition coincides with the downturn, so the new leadership's first real test is a defensive one, and the market questions whether the second generation can manage a cycle as deftly as the founder did.