Worldex Industry & Trading Co., Ltd. (KOSDAQ: 101160) - Deep Dive

A Korean maker of the consumable silicon and quartz parts that get eaten away inside plasma etch chambers. Every time Samsung or SK Hynix etches a wafer, a little bit of Worldex's product is sacrificed - and has to be reordered.

A note on sources and cadence: Worldex has a December fiscal year-end and reports quarterly under Korean rules. As of today (2026-07-04) the most recent released period is Q1 2026 (ended 2026-03-31, filed mid-May 2026). The next release, H1/Q2 2026 (ended 2026-06-30), is scheduled for around 2026-08-14 and has not yet been published. Unlike large-cap Korean names, Worldex does not hold quarterly investor conference calls or publish transcripts. Where the brief asks for "six concalls," I use the company's last six reporting periods (Q1 2026, FY2025, Q3 2025, H1 2025, Q1 2025, FY2024) together with management's public IR statements - the CEO's 2026 New Year message, the value-up plan announcement, and the June 2026 extraordinary general meeting (EGM) - as the equivalent record. This is stated plainly so the reader knows exactly what evidence underpins Sections 7 and 9.

1. What the company does

Worldex makes the parts that plasma etching machines slowly destroy while they carve circuit patterns into silicon wafers.

To understand the business you have to understand one physical fact about chipmaking. To build a chip, you deposit thin films on a wafer and then selectively remove material by bombarding it with plasma - a superheated, chemically aggressive soup of ionised gas - inside a sealed vacuum chamber. That plasma does not politely stop at the wafer. It attacks everything around it, including the fixtures that hold and shape it: the electrode above the wafer, the ring that surrounds it, the shower plate that distributes gas. Those fixtures are made of ultra-pure silicon and quartz precisely because silicon and quartz are chemically similar to the wafer and won't contaminate it. They are consumable. A silicon focus ring, for example, needs replacing after roughly 200 wafers. The chamber part gets sacrificed so the wafer stays clean.

Worldex makes those sacrificial parts. Its core products are the silicon electrode, silicon ring, and silicon boat used in etch chambers, plus quartz rings, tubes and boats, fine-ceramic parts (alumina), and physical-vapor-deposition (PVD) target materials. Because they wear out on a fixed schedule tied to wafer volume, demand is genuinely recurring: this is a razor-and-blade business where the fab is the razor and Worldex sells blades. When Samsung and SK Hynix run their fabs harder to feed the AI memory boom, they burn through etch parts faster, and Worldex ships more.

The founding story explains why the company matters in Korea. Worldex was established in 2000 (some records say the group's roots go back to the late 1990s as a trading house, hence "Industry & Trading" in the name). In 2001 it developed the micro-hole machining technology for silicon electrodes in-house. This mattered enormously because until then Korea imported these parts entirely from Japan. Worldex was the first Korean company to localize (국산화) the silicon electrode and silicon ring. In 2009 it extended the same localization playbook to high-purity quartz etch parts. That same year it bought its US silicon-ingot supplier outright (see Section 2), turning a supply relationship into vertical integration.

Management frames the whole enterprise as a "total solution" in etch-chamber materials - from raw ingot to finished, cleaned part - built across silicon, fine ceramics, quartz, and new materials such as silicon carbide (SiC) and aluminum nitride (AlN). The strategic ambition, repeated in the CEO's 2026 New Year message, is to move from a Korean localization champion to a "global tier-one" parts supplier.

Concretely, here is what Worldex does for a customer. A chipmaker (or the equipment maker that serves it, such as Tokyo Electron or SEMES) specifies an etch part to a tight purity, dimensional and surface-finish standard. Worldex sources or grows the silicon ingot (often through its US subsidiary), machines it into the electrode or ring with the micro-holes and geometry the plasma recipe requires, cleans it to remove particle and metal contamination, qualifies it against the customer's process, and ships it. When that part wears out weeks later, the customer reorders. The relationship is sticky because a mis-specified or dirty part can crater wafer yield - so once you are qualified into a process, the customer does not casually switch.

2. Business segments

Worldex reports two product families and runs a set of overseas subsidiaries that sit inside the supply chain rather than operating as independent businesses.

Silicon parts (~72-73% of revenue)

This is the core and the origin of the company. Silicon parts are the electrodes, rings, boats, and shower/gas plates made from ultra-pure silicon that sit inside plasma etch chambers. These are the highest-value parts in the etch consumable set because they directly shape the plasma field and therefore the etch profile - they "determine the etching capability," in the industry's phrase. Silicon has grown from under 60% of Worldex's mix in 2022 to roughly 72-73% now, meaning the company has become more concentrated in its strongest, most differentiated line as the memory-parts cycle turned up.

The core capability here is the full in-house chain: ingot supply, precision machining of features like the micro-holes in an electrode, and the cleaning/qualification step that removes contamination. That chain took two decades to build and is why the company can claim "material-through-to-part" integrated production. This is the margin and identity engine of the group, and management talks about it as the platform onto which everything else is bolted.

Competitively, silicon parts put Worldex up against WONIK QnC and Hana Materials in Korea, and against Japanese incumbents such as Mitsubishi Materials globally. Worldex's edge is its long-qualified position at Samsung and SK Hynix and its ownership of upstream ingot supply; its exposure is that Hana Materials is substantially larger and is pushing hard into SiC rings, and WONIK QnC is a much bigger, more diversified parts house.

Quartz and other parts (~28% of revenue)

Quartz parts (rings, tubes, boats) are used where extreme thermal and plasma resistance is needed - high-purity fused quartz withstands the high-temperature, high-plasma-density environment inside etch and deposition equipment. Worldex localized quartz etch parts starting in 2009. This bucket also carries the company's fine ceramics (alumina), its PVD target materials (used in display and semiconductor sputtering/deposition), and its newer materials development in SiC and AlN.

Quartz exists as a separate line because the material science, forming and machining are different from silicon, and because it serves partly different chamber positions and process steps. This segment is the diversification and optionality bet: it is where the SiC focus-ring opportunity and the ceramic-parts expansion live, both of which management has flagged as future ASP and volume drivers. Here the sharpest competitor is TCK (TC&I), the dominant Korean synthetic-quartz and SiC parts specialist, alongside WONIK QnC's quartz business and SK Enpulse (Solmics) in ceramics.

The subsidiary network (supply chain, not a segment)

Worldex consolidates four overseas subsidiaries that function as its supply and distribution spine rather than standalone segments:

• West Coast Quartz Corporation (WCQ), USA (100% owned): acquired in 2009 for about KRW 34.8 billion (then ~90% of Worldex's equity - a bet-the-company move). WCQ is a California silicon-ingot and components operation. It made Worldex vertically integrated into its own raw material. Note the important wrinkle: WCQ has been running operating losses that continue to drag group profitability, disclosed as a limiting factor on margins as recently as Q1 2026. The upstream integration is strategically valuable (control of ingot quality) but is not currently a profit center.
• WCQ Taiwan, WCQ Japan (est. 2007), WCQ Asia-Pacific / Singapore (est. 2003): regional sales and component-distribution arms that put Worldex close to non-Korean customers (TSMC in Taiwan; Kioxia/Sony in Japan; broader Asia-Pacific accounts). They are why roughly two-thirds of revenue is export.

3. Products and business detail

Silicon electrode. The plate that sits above the wafer and shapes the plasma, drilled with the micro-hole pattern that distributes etch gas. This is Worldex's flagship and the product it localized in 2001. The hard part is the micro-hole machining to spec plus the purity/cleanliness that prevents metal contamination of the wafer.

Silicon ring / focus ring / confinement ring / edge ring. The rings that surround the wafer edge and confine and focus the plasma so the etch is uniform to the wafer's edge. These are the highest-turnover consumable - replaced roughly every ~200 wafers - which is what makes the business recurring. The industry is migrating some ring positions from silicon and quartz toward CVD-SiC because SiC lasts longer and resists plasma better; this is both an opportunity (higher ASP) and a threat (if a competitor owns the SiC ring, Worldex loses the socket).

Silicon boat / wafer boat. Carriers that hold multiple wafers during batch processing.

Shower plate / gas shower plate (GSP). Distributes process gas evenly across the chamber; named by management as one of the integrated parts Worldex machines from Si/SiC/alumina/AlN.

Quartz parts (rings, tubes, boats). High-purity fused-quartz fixtures for the highest-temperature chamber positions.

Fine ceramics (alumina, Al₂O₃) and new materials (SiC, AlN). The forward-looking material set. Worldex says it has an integrated production system spanning Si, SiC, alumina and AlN, machining each into electrodes and shower plates. AlN is prized for thermal conductivity; SiC for plasma durability.

PVD target materials. Sputtering targets used to deposit thin films in display and semiconductor processes.

LED lighting (subsidiary, minor). A legacy non-core line carried through a subsidiary; immaterial to the investment case.

Manufacturing and geography. The production base is in Korea, centered on Gumi (Gyeongsangbuk-do), including a new-facility expansion in the Gumi 5th Industrial Complex begun in 2022 that lifted capacity and underpinned several years of high-single-digit average growth. Upstream ingot capability sits at WCQ in California. Distribution runs through the Taiwan, Japan and Singapore subsidiaries. The defining process attribute is the integrated "material-to-part-to-cleaning" chain: Worldex grows/sources ingot, machines the part, and cleans/qualifies it in-house, which is both a cost-control and a quality-control lever versus pure machining shops.

Milestones that changed the business: 2001 in-house silicon-electrode micro-hole technology (first Korean localization); 2008 KOSDAQ IPO; 2009 quartz-parts localization and the WCQ acquisition (vertical integration); 2022 Gumi 5th-complex expansion (capacity step-up); 2025-2026 the KRW 260 billion capex-and-diversification plan into SiC/ceramics and adjacencies.

4. Customers

Worldex sells to the memory and logic chipmakers and, indirectly, to the etch-equipment makers that serve them. Named customers across filings and press include Samsung Electronics, SK Hynix, Micron, TSMC, Kioxia (formerly Toshiba), Intel, and Sony, plus the equipment houses Tokyo Electron and SEMES (Samsung's captive tool maker).

The buying split is instructive. Samsung Electronics and SK Hynix together accounted for roughly 45% of Q3 2025 sales, and domestic sales were about 34% of revenue - meaning the other ~66% is export, flowing largely through agents and the overseas subsidiaries to Micron, Kioxia, Intel and others. So the customer base is concentrated at the top (two Korean memory giants) but genuinely diversified geographically.

Who decides and on what basis: the buyer is a fab's process-integration and equipment-engineering function, sometimes routed through the OEM equipment maker. The criteria are, in order, process qualification (does the part hold yield?), purity/particle performance, dimensional consistency, delivery reliability, and price. The sales cycle to get a new part qualified into an advanced process is long - months of testing - because a bad part shows up as scrapped wafers, the most expensive failure in the building.

That qualification is the switching cost. Once a Worldex electrode or ring is designed into a specific etch recipe and has demonstrated yield, swapping to another vendor means re-qualifying, which fabs avoid unless forced by price, capacity or a supply scare. This is why a 20-plus-year incumbent position at Samsung and SK Hynix is the company's most valuable asset - it is an installed base defended by requalification friction, not a contract. Business is largely recurring reorder flow tied to wafer volume rather than long-dated take-or-pay contracts, which makes revenue track fab utilization closely: when memory makers throttle back, parts orders fall within a quarter (visible in the softer Q1 2026 top line as front-end utilization wobbled), and when they run hot for HBM and high-layer NAND, orders surge.

5. Competitive landscape

The Korean etch-consumables space is a cluster of specialists, each anchored in a material or a chamber position, all selling into the same two or three giant customers. Worldex is a mid-sized player by market value - roughly KRW 0.45-0.5 trillion - sitting below several larger peers.

• WONIK QnC (KOSDAQ 074600): the broadest Korean quartz-and-parts house, also making silicon electrodes and rings; part of the WONIK group, much larger and more diversified than Worldex (2026 revenue guided around KRW 1 trillion). Worldex's disadvantage is scale and breadth; its answer is depth in silicon and its integrated ingot supply.
• Hana Materials (KOSDAQ 166090): the closest direct competitor in silicon etch parts (electrodes, rings) and an aggressive mover into SiC rings; larger than Worldex and growing fast on the memory recovery. This is the peer to watch on the SiC transition.
• TCK / TC&I (KOSDAQ 064760): the high-margin Korean specialist in synthetic quartz and CVD-SiC consumables, with a strong share in SiC rings. Overlaps Worldex most in quartz and in the SiC-ring opportunity Worldex wants to enter.
• KNJ (K&J, 케이엔제이): a smaller specialist pushing CVD-SiC focus rings; a direct threat in the specific socket the whole industry is migrating toward.
• SK Enpulse / Solmics: ceramics and parts arm of SK, a competitor as Worldex expands into fine ceramics (Worldex has publicly shown interest in the SK Enpulse fine-ceramics space).
• Mitsubishi Materials and other Japanese incumbents: the original suppliers Worldex displaced through localization; still relevant globally, especially for advanced silicon parts.

(Market caps are approximate peer-size references as of mid-2026 and move with the market; they are not valuation benchmarks for Worldex.)

The barriers to entry are real but not impregnable. On the positive side: customer qualification is slow and yield-critical, purity and cleaning process knowledge is hard-won, and a 20-year incumbent relationship is defended by requalification friction. On the exposed side: this is still a machined-and-cleaned commodity part at heart, several well-capitalized Korean rivals sell into the identical customers, and the single most important structural shift - the migration of ring positions from silicon/quartz to CVD-SiC - is being led by rivals (Hana Materials, TCK, KNJ) as much as by Worldex. If Worldex is slow to win SiC-ring sockets, it risks defending a shrinking silicon-and-quartz base while the value migrates. The honest read: Worldex has a durable niche and a genuine incumbency moat at its two anchor customers, but it is a mid-tier player in a competitive, partly commoditized field, not a runaway category leader.

6. Industry

Demand for Worldex's products is driven by one thing above all: how many wafers the world's fabs are etching, and how many etch steps each wafer needs. Two secular forces push both up. First, 3D NAND vertical scaling - stacking ever more layers (well past 200) - requires deeper, harder high-aspect-ratio etches, which consume chamber parts faster and demand better materials. Second, the AI/HBM boom is running memory fabs hot and adding advanced-packaging and DRAM capacity, directly lifting etch-parts burn. Worldex's own Q3 2025 strength was explicitly attributed to "the HBM boom's trickle-down effect."

Market size: the silicon rings and silicon electrodes for etching market was roughly US$1.7 billion in 2025, projected to reach about US$2.8 billion by 2031 at a ~7.3% CAGR (Valuates/QYResearch). The narrower silicon-ring market is put at ~US$0.9 billion (2024) growing to ~US$1.3 billion (2030) at ~6.3%. The fastest-growing adjacent pocket is CVD-SiC focus rings, forecast at ~10.4% CAGR through 2033, as SiC displaces quartz in high-wear positions because it lasts up to ~30% longer. So the base market grows steadily mid-single-to-high-single digits, but the value is migrating toward SiC.

Where Worldex sits in the chain: it is a Tier-2/consumable-materials supplier feeding both the chipmakers directly and the etch-equipment OEMs. Its localization role matters for supply-chain security: Korea historically imported these parts from Japan, and government and customer preference for domestic, requalified suppliers (sharpened by the 2019 Japan-Korea export-control episode) is a structural tailwind for Korean parts makers generally.

Regulation and certification are customer-driven rather than statutory - the "approval" that matters is process qualification at each fab, not a government license. Cyclicality is high and tied to the memory capex/utilization cycle: parts orders track wafer starts with roughly a one-quarter lag, so Worldex feels both the downturns (softer 2023 and a wobble in early 2026) and the up-cycles (HBM-driven 2024-2025) quickly. The industry tailwinds are AI memory intensity, NAND layer-count growth, and Korean localization preference; the headwinds are memory-cycle volatility and the risk of being on the wrong side of the SiC transition.

7. Growth triggers

Drawn from Worldex's reporting-period disclosures and public IR statements over the past ~18 months (the company does not hold earnings calls, so these are sourced to filings, the CEO's New Year message, and the value-up announcement, dated accordingly).

• KRW 260 billion investment program through 2030 (value-up announcement, mid-2025; reaffirmed FY2025 report, Mar 2026). KRW 190 billion into silicon/quartz/ceramic parts capacity and facilities, KRW 20 billion into next-generation process-material R&D, and KRW 50 billion into M&A/equity stakes in robotics, battery, defense, and thermal-management adjacencies.

Management framed cash reserves as "a core asset for responding to market uncertainties and securing next-generation technologies," positioning the capex as a deliberate choice of growth over near-term payout.

• Gumi 5th-complex capacity expansion (ongoing; referenced FY2024 and FY2025 reports). Management expects the current expansion to reproduce the growth effect of the 2022 investment, which underpinned ~9% average annual growth over the prior five years.

• New-material entry: SiC and AlN parts (FY2025 report, Mar 2026; CEO New Year message, Jan 2026). Building an integrated Si/SiC/alumina/AlN production system to supply electrodes and shower plates - the direct route into the higher-ASP CVD-SiC ring transition the industry is undergoing.

• HBM/AI-driven memory-parts demand (Q3 2025 and FY2025 reports). Management tied Q3 2025 profit strength to HBM ramp at Samsung and SK Hynix, and positioned the company for continued AI-server and advanced-packaging pull.

The CEO's 2026 slogan - "empower autonomous leadership to drive a new paradigm" - was paired with an explicit aim to reach "global tier-one" status by advancing etch-material processes and building "overwhelming quality and price competitiveness" in AI and automotive semiconductors.

• Adjacency M&A optionality (value-up announcement, 2025). The KRW 50 billion earmarked for robotics, battery, defense and thermal-management stakes is a stated, funded intent to diversify beyond semiconductor consumables, though no specific target has been named.

• Diversification into fine ceramics (press/IR, May 2025). Worldex has publicly signaled interest in the fine-ceramics space (the SK Enpulse adjacency), consistent with its alumina/AlN build-out.

8. Key risks

WCQ subsidiary losses drag group profit. The US ingot subsidiary that makes Worldex vertically integrated has been running operating losses that the company itself cites as a limiter on profitability, most recently in the Q1 2026 report. The mechanism: a structurally loss-making upstream unit consumes the margin the Korean parts business earns. This is a high-probability, moderate, ongoing drag rather than a catastrophe - but until WCQ turns, group returns are capped below what the core Korean operation would earn alone.

Memory-cycle whiplash. Because orders track fab utilization with a short lag and two Korean memory makers are ~45% of sales, a memory downturn hits Worldex within a quarter. Q1 2026 already showed a softer top line "as front-end device-maker utilization fluctuated." High probability, moderate-to-high impact, and largely outside management's control.

Losing the SiC-ring transition. The single most dangerous structural risk. The industry is migrating high-wear ring positions from silicon/quartz to CVD-SiC, and rivals Hana Materials, TCK and KNJ are leading it. If Worldex's SiC/AlN build-out is late or sub-scale, it defends an eroding silicon-and-quartz base while the ASP growth accrues to competitors. Medium probability, high long-term impact - this is the risk that could quietly hollow out the franchise over five years.

Customer concentration. Samsung and SK Hynix are both the crown jewel (a defended incumbency) and the concentration risk. A qualification loss, an in-sourcing decision, or a pricing squeeze from either would be hard to replace given how much of the base they represent.

Governance and capital-allocation risk - now an active dispute. Founder-CEO Bae Jong-sik owns ~34.8% and three of four inside directors are family members. The company's three-year average dividend payout was ~2.3%, and over that window total dividends (~KRW 3.6 billion) were smaller than the CEO's cumulative compensation (~KRW 4 billion). Second-largest shareholder VIP Asset Management (15.6%) went activist in June 2026 and won a decisive EGM vote against the board's director-pay proposals.

VIP publicly criticized the board's KRW 260 billion investment plan and 10% payout target as "lacking substance," arguing the company's ~KRW 230 billion cash pile and ~KRW 160 billion of three-year operating cash flow could support far more aggressive returns.

The mechanism of risk here runs two ways: entrenched family control could keep capital allocation shareholder-unfriendly (value trap), or the escalating activist pressure could force change (catalyst). Either way, it is now a live, material factor rather than a background concern.

9. Walk the talk

The six reporting periods used: FY2024 (filed Mar 2025), Q1 2025 (May 2025), H1 2025 (Aug 2025), Q3 2025 (Nov 2025), FY2025 (Mar 2026), and Q1 2026 (May 2026), supplemented by the mid-2025 value-up announcement, the Jan 2026 CEO message, and the Jun 2026 EGM. Because Worldex publishes no call transcripts, this assessment leans on filed results versus stated intentions and on the public IR record.

On the operational promises, management has been broadly credible. Through FY2024 and into 2025 the company positioned itself to benefit from the HBM/AI memory ramp, and the Q3 2025 result delivered on that thesis - operating profit rose sharply on HBM-driven demand, exactly the mechanism management had pointed to. The multi-year narrative that the 2022 Gumi expansion would drive high-single-digit growth is consistent with the actual ~9% five-year average growth the company cites. On the core "we are qualified, incumbent, and levered to memory intensity" story, management's words and the outcomes line up. The one persistent operational gap it has been honest about, rather than hiding, is WCQ: the loss-making US subsidiary has been repeatedly disclosed as a profit limiter, most recently in Q1 2026. That is a promise not yet delivered (a turnaround) but transparently flagged, not buried.

On capital allocation, the record is where credibility frays, and the sequence tells the story. For years the company retained almost all its earnings - a ~2.3% payout - while building cash. When challenged, management's response in mid-2025 was the value-up package:

"enhance corporate value through predictable shareholder return policies and continuous growth investments" - a KRW 260 billion investment plan paired with a promise to lift the payout ratio toward a 10% average.

The second-largest shareholder judged that promise inadequate and said so publicly. The board then attempted to raise its own director-compensation ceiling (from KRW 7 billion to KRW 8 billion), with early versions reportedly including golden-parachute provisions - at the same moment it was asking shareholders to accept a still-modest payout. Independent and foreign shareholders, including Norway's sovereign wealth fund, sided with the activist: at the June 29, 2026 EGM the board's pay proposals were voted down, one by a ~95% margin among non-controlling shareholders.

The plain assessment: on operations, this is management that does roughly what it says - honest about the WCQ drag, and correct about the memory-parts up-cycle. On capital allocation and governance, the pattern before mid-2026 was one of retaining cash and rewarding insiders while returning little to shareholders, and the value-up plan reads as a reactive, minimum-viable concession rather than a conviction shift. The 2026 proxy defeat means the talk on returns is now being audited in real time by shareholders who no longer take management at its word. Watch whether the 10% payout and any buyback actually materialize in FY2026 - that is the promise that will settle the question.

10. Shareholder friendliness index

Dividends. Worldex has been a persistent under-distributor. Its three-year average dividend payout ratio was about 2.3%: on roughly KRW 160 billion of cumulative net profit, it paid only ~KRW 3.6 billion in total dividends. The declared 2025 dividend was 100 won per share (about KRW 1.6-1.7 billion in aggregate on ~16.5 million shares). Per-share dividends in the prior two years were similarly token; I could not verify an exact, differing DPS for each of 2023 and 2024 beyond that the aggregate remained minimal and the payout stayed near the low-single-digit percent range. The telling comparison, surfaced by activist VIP Asset Management and Korean press, is that total three-year dividends were smaller than the CEO's ~KRW 4 billion of cumulative pay - a rare and unflattering signal.

Buybacks and dilution. Over the last three years Worldex executed no meaningful share buyback or cancellation program - the absence of buybacks is precisely the core of the activist campaign. Shares outstanding sit around 16.5 million and have been broadly flat (no large issuance and no retirement), so there is neither buyback-driven shrinkage nor material option dilution. In mid-2025 the company announced a value-up plan promising to raise the payout ratio toward a 10% average over three years, and second-largest holder VIP has demanded at least KRW 20 billion of annual buyback-and-cancellation plus a 40%+ net-profit payout from 2026 onward; as of the latest reporting, those are demands and promises, not executed returns. (Note: the last-90-day window shows no buyback either; the three-year absence above is from annual-report and exchange-announcement history, not just the recent window.)

Verdict: Hoards Capital - a founder-controlled balance sheet that retained nearly all earnings and out-paid its shareholders to its own CEO, now under real activist pressure that may (but has not yet) force change.

11. Insider activities

The dominant, and most important, insider/large-holder activity of the last 12 months is the accumulation and activism of the second-largest shareholder, disclosed through Korea's 5%-rule (대량보유) reports on DART.

Reading the buys. VIP Asset Management steadily increased its position through late 2025 and into June 2026, then converted its filing purpose to active engagement. This is not routine passive accumulation - it is a value-oriented institution building a blocking position and then contesting the board. The signal it sends is about perceived undervaluation and trapped capital, not about near-term operating momentum: a sophisticated domestic manager judged the shares cheap enough, and the balance sheet lazy enough, to justify its first public proxy campaign in more than two decades. That is a constructive/bullish signal on the capital-allocation catalyst, distinct from an insider-conviction buy on the business itself.

The controlling insider. Founder-CEO Bae Jong-sik holds ~34.8%, with three of four inside directors being family members. Within the search budget I did not find evidence of open-market purchases or sales by Bae or other officers over the last 12 months (the family stake appears stable); the controlling family's posture has been defensive - trying to raise director compensation - rather than buying or selling stock.

Net assessment. Insider/large-holder activity over the past year is dominated by one actor: an activist institution buying and pressing, against a static founder-controlled core. There is no cluster of management open-market buying that would signal insider conviction in the stock, and no material insider selling that would signal alarm. The meaningful signal is the activist accumulation and its EGM win - a governance catalyst rather than a fundamental one. Net read: mildly bullish on the catalyst, neutral on insider conviction in the business - the people buying are outside shareholders forcing change, not managers backing their own company with their own money. Beyond the DART 5%-rule filings summarized here, granular director-level 임원·주요주주 transaction data was not located within the search budget.

12. Scenarios

Bull case. The AI-memory super-cycle runs for years, and Samsung and SK Hynix keep their fabs pinned near full utilization to feed HBM and high-layer NAND, burning through Worldex's electrodes and rings on a fixed schedule. The Gumi expansion fills, and - critically - Worldex converts its SiC/AlN build-out into qualified CVD-SiC ring sockets at its anchor customers, so it participates in the high-ASP material transition instead of watching rivals take it. The WCQ subsidiary finally turns, removing the profit drag and letting the group's true parts-margin show. Meanwhile the activist campaign works: management, chastened by the EGM defeat, actually executes the buyback-and-cancellation and lifts the payout, and the market re-rates a formerly capital-hoarding small-cap into a shareholder-friendly compounder with a defended niche. Diversification stakes in robotics or thermal management add an option no one is paying for today.

Base case. Worldex remains a solid, incumbent Tier-2 etch-consumables supplier whose results breathe with the memory cycle - strong in HBM-heavy quarters, softer when front-end utilization dips, as it did in early 2026. The Gumi capacity supports steady high-single-digit growth over time. The SiC transition happens gradually and Worldex holds some, not all, of the new sockets, keeping it relevant but not dominant against Hana Materials and TCK. WCQ stays a modest drag or slowly improves. On capital, the value-up plan is delivered in its minimal form - payout drifts toward 10%, a token buyback appears - enough to keep the activist engaged but not transformed. The company stays what it is: a competent niche parts maker with an incumbency moat, a governance discount, and a memory-cyclical earnings stream.

Bear case. The memory cycle rolls over, utilization at Samsung and SK Hynix falls, and parts orders drop within a quarter, compounding the WCQ losses into a genuinely weak stretch. Simultaneously, the CVD-SiC ring transition accelerates and Worldex is late - Hana Materials, TCK and KNJ lock up the SiC sockets, and Worldex is left defending a structurally shrinking silicon-and-quartz base. Customer concentration bites: an in-sourcing or dual-sourcing decision at one anchor account removes a slug of revenue that cannot be replaced. And on governance, the founder family digs in, the value-up promises quietly under-deliver, the activist eventually tires or exits, and the stock stays a cash-rich value trap where the capital never reaches shareholders. The moat holds enough to keep the company alive, but the value migrates around it.