SCREEN Holdings Deep Value Investment Analysis

Hunting ten-year five-baggers among great growth stocks — pressing the upside question: "Can it get much bigger?"

• QHow high is its market ceiling? Is it expanding an existing pie, or creating an entirely new market?4/10

  A

  Bottom line: SCREEN is expanding an existing pie that is not especially large, rather than creating a new market. For Baillie Gifford’s “fivefold in ten years” narrative, the duration of its runway is acceptable, but the breadth of that runway, meaning the absolute size and slope of TAM, is clearly narrow. Through the Baillie Gifford lens, this is only medium-to-weak overall. Semiconductor cleaning equipment is a mature process step that has existed for decades. SCREEN is the global leader in this niche, and its growth comes from making the existing pie thicker as fab capital spending and process upgrades advance, not from opening up demand that previously did not exist. This is fundamentally different from the blue-sky narrative Baillie Gifford prefers: defining and dominating a new category whose TAM can expand 10 times.

  Start with the market ceiling’s absolute height, which is the biggest hard constraint. SCREEN’s core battlefield is wafer wet-cleaning/surface-treatment equipment, and third-party estimates put the global wafer cleaning equipment market at only about $6.4 billion in 2025, about $6.9 billion in 2026, and about $9.9 billion in 2031, with a CAGR of about 7.5% (Mordor Intelligence). Put that figure into the broader semiconductor equipment landscape and its narrowness is clear: according to SEMI’s December 2025 figures, global semiconductor equipment sales are about $133.0 billion in 2025, about $145.0 billion in 2026, and are expected to reach a record $156.0 billion in 2027, while wafer fabrication equipment (WFE) is about $115.7 billion in 2025, rising 9% in 2026 and 7.3% in 2027 to about $135.2 billion (SEMI / Tom's Hardware). In other words, the cleaning-equipment TAM where SCREEN mainly competes is only about 5% of total WFE. It performs an indispensable but relatively small-value cleaning step inside a much larger machine dominated by the “big three” of lithography/etch/thin film. For a company already valued at about ¥2.54 trillion, or roughly a $17.5 billion scale, relying on a core TAM of $6.0–7.0 billion to support a “fivefold in ten years” outcome, meaning a market value above ¥12 trillion, is mathematically very hard if the only lever is thickening this pie.

  Next, look at the runway’s duration and slope. This is where SCREEN looks relatively respectable, but it is still an existing pie getting larger, not a new pie. Cleaning equipment has real structural drivers for long-term growth: as process nodes move toward 2nm/1.6nm, sub-10nm particle removal becomes mandatory, and the number of cleaning steps increases monotonically as nodes shrink; HBM, advanced logic, advanced packaging, and chiplets all raise the importance of surface treatment. The report accordingly argues that “advanced processes/advanced packaging have increased the importance of cleaning,” and that the core moat has “slightly widened.” SEMI likewise notes that 300mm fab equipment spending will rise 18% in 2026 to about $133.0 billion and 14% in 2027 to about $151.0 billion, driven by AI-led advanced logic, HBM memory, and capacity self-sufficiency (SEMI 300mm report). But note carefully: this is an industry-level expansion of an existing pie. A slope of about 7%–9% CAGR is a respectable long-term growth industry, far from exponential new-market creation. It is also highly cyclical. In the report, FY2026 revenue is already down -3.1% year on year and operating profit is down -9.7%, which proves this curve is not linearly upward; when the cycle turns, the pie shrinks. Baillie Gifford’s firepower is concentrated on upside imagination in years 3–10, but for cleaning equipment, years 3–10 look more like “steadily thickening with fab capex and Moore’s Law” than “demand going from zero to mass adoption.”

  Third, examine SCREEN’s penetration and share within this pie, which paradoxically lowers the ceiling. In 2024, SCREEN ranked global number one in single-wafer cleaning, batch cleaning, and spin scrubber cleaning equipment (the report cites Gartner), with single-wafer cleaning share already high at about 42%–45% (SCREEN SWOT/ICS figures). The top five cleaning-equipment vendors, SCREEN, Tokyo Electron, Applied Materials, ACM Research, and Lam, together account for about 65% of revenue (Mordor Intelligence). This is a double-edged sword: it proves the moat is real, but also means SCREEN is already close to a share ceiling in its strongest niche. It is hard to multiply revenue several times by taking more share; upside is mainly constrained by the TAM growth rate of the industry itself, about 7%–8% per year. This is the opposite of Baillie Gifford’s preferred setup of low penetration and large room for share expansion. An ideal LTGG company often storms through a huge new market with single-digit penetration; SCREEN is defending a small, mature pool where it already has a 40%–50% share.

  The qualitative call on “expanding an existing pie vs creating a new market” is clear: it is the former. Wet cleaning has been a standard semiconductor manufacturing step for decades. SCREEN traces its origins to screen printing and photomasks; it did not create a new demand category from thin air. What it does is make an existing process step harder, more expensive, and more numerous per wafer as nodes upgrade, thereby increasing the cleaning-equipment value per wafer. That is a textbook case of deepening an existing pie, not building a new one from scratch. It is trying to extend into adjacent process steps along advanced packaging and surface treatment, but these remain adjacent expansions within the existing WFE landscape, not a “new market” in the Baillie Gifford sense. The report’s framing, that the group has 83% of revenue in SPE and is “an excellent niche leader in a good industry, but not an invincible monopolist,” is consistent with this judgment.

  Overall assessment for the Baillie Gifford “fivefold in ten years” narrative: the ceiling is long enough but not wide enough, creating a material constraint. Plainly: (1) the core TAM is only $6.0–7.0 billion, with a CAGR of about 7.5%, so both absolute size and slope are too small to support exponential growth for a single stock; (2) SCREEN already has a 40%–50% share in its strongest niche, leaving little room for share expansion, and future growth is highly dependent on industry capex and a demand pool that is itself strongly cyclical; (3) this thickens an existing pie rather than creating a new market, lacking the combination Baillie Gifford values most: a new category, low penetration, and a long, snow-rich slope. To deliver “fivefold in ten years,” SCREEN must simultaneously get a favorable sequence right: AI capex stays strong for a long period, cleaning-step value keeps rising with node migration, share is not eroded by Chinese local substitution, and valuation keeps a growth premium. It is not naturally carried by the width of the ceiling itself. Through Baillie Gifford’s growth lens, I rate this “market ceiling” question weak to medium: the industry runway is real, but the slope is not steep enough, the pool is not large enough, and the company is already near its share limit. SCREEN is an excellent cyclical growth leader, but not the kind of great growth stock whose ceiling is invisible in the Baillie Gifford sense.

  Jun 5, 2026
• QCan its revenue at least double over the next five years? Will growth mainly be driven by volume, price, or new businesses?4/10

  A

  Bottom line: doubling revenue over five years, to FY2031, to about ¥1.2tn is not the base case and is closer to the optimistic ceiling. It would require about a 15% compound annual rate, while the company itself places “¥1tn sales” further out, ten years later in FY2033, ending March 2033. In other words, the official roadmap itself does not treat “doubling in five years” as the target. What can support the doubling narrative is the strong rebound starting in FY2027 and the AI equipment supercycle. What holds it back is its strong cyclicality, its cleaning share already near the ceiling, and its 42% China exposure. By source of growth, it is almost entirely volume, meaning more process-equipment shipments and new fab expansions; price is a soft support, and new business, advanced packaging, is an accelerator rather than a fresh engine from scratch. Through Baillie Gifford’s “5 times over the next ten years” lens, SCREEN answers this question as “qualified but not impressive”: it can explain upside elasticity, but management’s own medium- to long-term slope does not match a five-year doubling.

  First, get the “now declining” premise right. This is a cyclical trough, not a trend reversal. The report’s FY2026 revenue of ¥605.7bn, down -3.1% year on year, is correct, but it is a one-year cyclical pullback. In its FY2026 results, the company simultaneously guided FY2027, ending March 2027, to net sales of ¥725bn, operating profit of ¥150bn, and net income of ¥110bn, implying revenue growth of about +19.7% and operating profit growth of about +22.4%, both record highs. In other words, the market, and the current share price, is not anchored on “recession”; it is anchored on “FY2026 bottoming, FY2027 record rebound.” This is exactly what the report repeatedly flags: the current price has already incorporated the FY2027 recovery expectation. Linearly extrapolating FY2026’s -3.1% as a five-year starting point would systematically understate growth; extrapolating FY2027’s +20% as normal would systematically overstate it. Both are wrong.

  The conflict between the arithmetic threshold for doubling and the company’s own slope is the point this question most needs to state honestly. Doubling from FY2026 ¥605.7bn to about ¥1.21tn in five years requires about 15% annualized growth. Yet the company’s medium- to long-term roadmap is net sales above ¥1tn and operating margin above 20% ten years later, in FY2033. From the FY2026 base, ¥1tn implies only about 5% annualized over ten years. In other words, the official vision of “reaching ¥1tn only after seven years” does not support “reaching ¥1.2tn in five years”. Even using FY2027’s ¥725bn as the springboard, reaching ¥1.2tn by FY2031 still requires about 13.6% annualized growth from FY2027 to FY2031. For an equipment-cycle business, that means “no meaningful down year for four years,” which is a demanding assumption in semiconductor capex history. The report’s neutral DCF assumes only 6% growth for the first five years, and the optimistic case only 10%; the gap versus the 15% needed for doubling precisely quantifies why the rating stops at Watch. The honest conclusion is: a five-year doubling is possible, but not the base case; it sits on the optimistic-scenario side of the report and depends heavily on the AI cycle staying intact.

  The external cycle provides a tailwind, not a rocket that guarantees doubling. SEMI’s latest December 2025 forecast: global semiconductor equipment sales are $133bn in 2025 (+13.7%), $145bn in 2026, and $156bn in 2027, breaking $150bn for the first time. Wafer fab equipment, the category most relevant to SCREEN, is expected to grow +9.0% in 2026 and +7.3% in 2027 to $135.2bn, driven by AI-led advanced logic, HBM/DRAM, and advanced packaging. The report’s body used an earlier 2026 figure of about $139bn; the latest figure has been revised up to $145bn, with the same direction and a higher scale. The key is the slope: WFE is growing at a high-single-digit rate over these two years, not expanding at a doubling pace. SCREEN’s FY2027 guidance of +20% outgrows the industry through share and product mix, but a parent market growing at a high-single-digit annual rate cannot easily support sustained company-level 15% compounding over the long term unless SCREEN keeps taking share. In its core cleaning equipment business it is already the global number one with 45%+ share, so the room to take share is naturally limited by the fact that it is already at the ceiling. This is where Baillie Gifford’s “market ceiling” question intersects with this one: parent-market growth plus a share ceiling together compress the probability of a five-year doubling.

  Growth breakdown: volume is the clear main driver, price is soft support, and new business is an accelerator. According to the FY2026 earnings call, management explicitly attributes FY2027 growth to three volume drivers: ① advanced logic/foundry capacity expansion, the largest engine, driven by AI infrastructure and memory shortages; ② a memory recovery, especially AI-driven DRAM/HBM and NAND bottoming; ③ advanced packaging as a newly integrated business, with investment staying high. Price, or price increases, is only a soft support. The report has already noted that FY2026 revenue fell 3.1% while operating margin still held at 20.2%, and SPE revenue fell 6.5% while operating profit still reached ¥122.7bn. That shows pricing power in key processes, but it is a defensive moat that cushions declines, not a lever for expanding revenue through price hikes. The CFO also credited better-than-expected profit to after-sales service/recurring revenue, meaning volume plus stickiness, rather than higher unit prices. Advanced packaging is the only new business credible enough to be called a “second growth pole,” but it remains an extension of the same wet-cleaning/coating capabilities. It extends the existing moat into new process nodes, not a new curve from scratch. It can add volume elasticity, but cannot support growth independent of the semiconductor cycle.

  There are three material drags on doubling, and they need to be on the table. First, strong cyclicality: FY2026’s -3.1% proves this business can genuinely decline. If the next five years include just one meaningful capex down year, 15% compounding largely breaks. The report’s “maximum loss risk” scenario, with the share price returning to ¥5,000–7,000 and a 45%–60% drawdown, corresponds exactly to “being re-rated as an ordinary cyclical stock rather than a growth stock.” Second, China exposure: China accounted for 42.4% of FY2025 revenue, and management expects it to fall to about 38% in FY2027, while warning that potential new regulations could hit investment in that market by 10–15%. Local substitution plus regulation is the tail risk most likely to cut a chunk out of volume over a five-year horizon. Third, the share ceiling: cleaning is already global number one at 45%+, and incremental contribution from taking share is diminishing. Future growth must rely more on passive volume increases from more single-wafer cleaning steps as nodes advance, not from conquering new territory.

  Honest score through the Baillie Gifford lens, for this question only: neutral to weak. SCREEN has two real ingredients for upside imagination: the AI equipment supercycle, and FY2027’s +20% rebound showing elasticity still exists. But the 15% compounding required for a “five-year doubling” is above both the parent-market WFE high-single-digit growth rate and the company’s own official vision of reaching only ¥1tn in ten years, or about 5% annualized. It is also constrained by cyclicality, China exposure, and the share ceiling. The conclusion remains as stated upfront: doubling is an optimistic scenario rather than the base case, and the source of growth clearly lies in volume, with price and new business secondary. This growth question is answered realistically and adequately, but not at the level Baillie Gifford seeks, where the slope is steep enough for the market to misjudge it. That is internally consistent with the report’s overall view: “a good company, but not at a good enough price.”

  Jun 5, 2026
• QFive years from now, what will take over as the next growth engine? Does this “second curve” exist today?3/10

  A

  Bottom line: SCREEN does not have a “second curve” today that can independently take over. Every candidate large and fast enough to carry the growth banner five years from now, advanced packaging cleaning, HBM cleaning, coater/developer, panel-level packaging coating/direct imaging, is essentially still a niche extension of the same main semiconductor wet-cleaning/surface-treatment curve. They rely on the same chemical-fluid management and particle-control know-how, sell to the same wafer/memory/packaging customers, rise and fall with the same semiconductor capex cycle, and are all included in the same SPE reporting segment, which accounted for 83.1% of group revenue in FY2025. The businesses truly independent of semiconductor capex, graphic arts printing GA, display coating FT, and PCB PE, together account for only about 16.5% of the group, grow modestly, and are being diluted in share. By scale, they cannot possibly “take over” from an SPE engine that still contributed ¥486.0bn in FY2026. So the honest answer is: the second curve does not exist today. SCREEN’s growth engine five years from now is most likely still today’s main curve itself, just applied to more advanced use cases. This is fully consistent with the report’s view that “buying SCREEN is essentially buying the wafer-cleaning equipment business” and “Watch: good company, but not at a good enough price.” It does not give you a new curve as an additional upside option.

  Below is a candidate-by-candidate screen, separating a true second curve from niche extensions of the main curve.

  ---

  ① Advanced packaging / HBM cleaning and surface treatment: the fastest-growing candidate, but it is an extension of the main curve, not a second curve. This is the key judgment.

  This is the most imaginative part of the SCREEN story, and the progress is real and verifiable: SCREEN SPE has joined Applied Materials’ EPIC Center as an innovation partner, integrating its “industry-leading cleaning, wet etch, and surface-treatment capabilities” with AMAT’s deposition/etch processes to build coordinated process solutions for advanced chips. The 180,000-square-foot Silicon Valley facility is planned to open in 2026; HBM/DRAM-related investment is guided to strengthen in the second half of FY2026. At the narrative level, AI/HBM/chiplets genuinely make pre- and post-packaging cleaning and surface cleanliness critical variables for yield.

  But to decide whether this is a “second curve,” three questions matter: scale, growth, and independence:

  • Independence = zero. Advanced packaging cleaning uses exactly the same wet-cleaning/surface-treatment core capabilities as SCREEN’s main business, sells to overlapping fab/memory/packaging customers, is driven by the same semiconductor capex cycle, and is financially housed in the SPE segment. It is not a new leg; the main leg has stepped into a new room. Once semiconductor capex enters a downturn, which this report lists as the top source of permanent capital loss, advanced packaging cleaning will not grow independently against the cycle. It rises and falls with the main curve. That is precisely not the “independent handoff” Baillie Gifford wants.
  • The scale is still very small. Public disclosures clearly indicate that SCREEN’s advanced packaging products are still a “billions of yen” small-scale business. Relative to SPE’s ¥486bn scale, even high growth over five years is unlikely to “take over.” The more realistic role is to add some slope to the main curve.
  • The underlying market growth needs to be separated carefully. The entire advanced packaging market rises from about $39.6bn in 2024 to $55.0bn in 2030, with CAGR of only 5.7%, which is not the explosive track many imagine. The truly fast market is hybrid bonding, where equipment revenue rises from about $152M in 2025 to about $397M in 2030, with CAGR around 21%, but the absolute pool is tiny, and the main beneficiaries are bonding tools/CMP/inspection, with cleaning only one link.

  Verdict: a high-quality enhancement of the main curve, not an independent second curve. It is a valid bull point, and this report indeed lists “advanced packaging increasing the importance of cleaning” as a moat-widening and core bull argument. But treating it as another leg that can independently carry growth overstates the company’s diversification.

  ---

  ② Coater/developer track: real incremental value, but also inside SPE, and SCREEN is the distant number 2 under a ceiling.

  This line itself is an extremely concentrated good market: every chip produced by TSMC/Samsung/Intel passes multiple times through TEL or SCREEN coater/developer equipment, and the two together have about 88% share. The problem is that the share distribution is extremely asymmetric: Tokyo Electron (TEL) holds about 89–90%, and is close to 100% monopoly in EUV / High-NA leading-edge tracks, while SCREEN is the distant number two blocked from the highest-value EUV nodes. ASML’s EUV/High-NA ramp mainly benefits TEL, not SCREEN.

  More importantly, coater/developer is also included in SPE and also depends on semiconductor capex, so independence still fails. It can provide SCREEN with a stable share, but it is not a new curve, and its growth ceiling is capped by TEL’s structural monopoly. Verdict: a stable branch inside the main curve, not a second curve, with a relatively low ceiling.

  ---

  ③ Panel-level packaging (PLP) coating/direct imaging: a new product line with the same roots as ①, still an SPE/FT extension.

  SCREEN’s Lemotia coating/drying system, aimed at FOPLP and glass-core substrates and expanded in 2025 to 300×300/310×310mm multiple sizes, and its LeVina direct imaging system, are concrete tools for selling picks and shovels into advanced packaging. Technically, SCREEN does have a leading position. But the qualitative judgment is the same as ①: small scale, same customers, same cycle, same SPE accounting. It is a product extension of the main curve, not an independent growth pole.

  ---

  ④ After-sales service / recurring revenue, including consumables: a cushion, not an engine.

  This report has confirmed that SPE after-sales service revenue growth helped improve profitability, and recurring businesses such as GA inks are also contributing. The direction is right, improving revenue predictability and smoothing cycles, but this line depends on the installed base and rises and falls with the same main curve. The report explicitly notes that “recurring revenue can only cushion; it cannot turn SCREEN into a software-like or medical-consumables-like high-predictability model.” Verdict: a stabilizer that improves quality, not a growth engine that can take over.

  ---

  ⑤ Graphic arts GA / display coating FT / PCB PE: these three are truly independent, but unfortunately they are small and not growing enough to take over.

  Only these three non-semiconductor segments are business-independent from semiconductor capex, which is their only “second-curve attribute.” But:

  • They are too small: FY2025 GA was ¥53.0bn (8.5%), FT ¥35.8bn (5.7%), and PE ¥14.1bn (2.3%). Together they were about ¥102.9bn, only about one-fifth of SPE’s ¥519.5bn. For them to “take over” from a ¥486bn-scale engine five years from now, they would need to multiply several times as a group, and there is no evidence supporting that.
  • Growth is mediocre, and their share is being diluted: GA is a mature printing-equipment business, including CTP/web inkjet/solder-mask direct imaging, with a structurally weak downstream printing market. FY2024 GA grew only about +4.7% year on year, more like a low-single-digit stable base than a growth-stock engine. As SPE expands rapidly, the group share of these three segments is declining passively. This report also states directly that “non-core business quality is below SPE” and that “whether non-core businesses continue to drag on group valuation quality” is a monitoring item.
  • They cannot rewrite the group narrative: the market prices SCREEN today as a high-cycle semiconductor growth stock entirely because of SPE; GA/FT/PE provide neither growth elasticity nor a second independent value anchor.

  Verdict: the only three truly independent segments are precisely too small and too slow to qualify as successors. They are ballast, not a second curve.

  ---

  Horizontal summary, one label each:

  Candidate	Scale	Growth	Independent of semiconductor capex?	Verdict
  Advanced packaging/HBM cleaning	Billions of yen (small)	Fast, but the underlying market is only ~5.7% CAGR; hybrid bonding is fast but the absolute pool is small	No (same capabilities/same customers/same cycle, included in SPE)	Main-curve enhancement, not a second curve
  Coater/developer	A piece inside SPE	Constrained by TEL’s near-100% EUV monopoly	No (included in SPE)	Main-curve branch, low ceiling
  Panel-level packaging PLP coating/direct imaging	Small	Early-stage	No (same-root extension)	Main-curve product extension
  After-sales/recurring revenue	Cushion-scale	Moderate growth	No (depends on installed base/same cycle)	Stabilizer, not an engine
  GA graphic arts	¥53.0bn / 8.5%	Low single digits, share being diluted	Yes	Truly independent but too small and slow to carry growth
  FT display coating	¥35.8bn / 5.7%	Weak cycle	Yes	Same as above
  PE (PCB)	¥14.1bn / 2.3%	Weak	Yes	Same as above

  Honest answer for the Baillie Gifford framework: the second curve does not exist today. SCREEN’s growth five years from now will almost certainly still be today’s main wafer wet-cleaning/surface-treatment curve, just pushed by AI/HBM/advanced packaging into more advanced and higher-value applications. This does not mean it lacks growth; it means its growth is highly single-threaded and tied to one point in the semiconductor capex cycle, with no backup leg that can independently take over when the main curve ebbs. For growth investors seeking upside imagination and years 3–10 firepower, this means upside depends on the same variable continuing to deliver, advanced-process/advanced-packaging capex not retreating and SCREEN maintaining cleaning share, rather than multiple curves stacking. That amplifies, rather than mitigates, the very risk this report worries about most: strong cyclicality plus a current price that has already overdrawn the upcycle.

  Jun 5, 2026
• QWhat is its core competitive advantage? Will this moat widen or narrow over the next three to five years?5/10

  A

  Bottom line: SCREEN’s moat is real but deep and narrow. It is the global number one in the cleaning step, not a broad-category leader in semiconductor equipment. Its core competitive advantage comes from three layers: ① global number-one share across the three cleaning subsegments, single-wafer/batch/spin scrubber, built on process qualification + PoR status + installed base that cannot be quickly replicated; ② switching costs created by yield risk, process introduction cycles, and stable operating records on customer production lines; ③ a global service network across the US, Europe, Korea, China, and Taiwan. Over the next 3–5 years, advanced logic/HBM/advanced packaging will raise the importance and unit value of cleaning, making the depth of this narrow moat deeper. But Chinese local substitution, with China representing 42.4% of revenue, and partial wet-to-dry substitution will keep pressuring the breadth from both sides. My net judgment: in the core high-end cleaning processes it retains, the moat slightly widens, with higher unit value and stickier customers; but if the view is the “serviceable scope” of the whole cleaning TAM, the moat’s boundary is being cut away by China and chipped away by dry processes. The net effect is roughly “deeper core, narrower edge,” close to neutral with a modest positive tilt, and absolutely not an invincible monopoly that keeps widening. It is a Buffett-style “real but limited moat,” not one that automatically widens with time. Details below.

  ---

  1. Sources of the moat: why “hard to copy” is real, but “hard to copy” does not mean “irreplaceable”

  I agree with the skeleton of the report’s moat judgment: switching costs are “relatively strong,” process/patent barriers are “strong,” scale advantages are “clear,” and network effects are “weak.” The key is to tie that to verifiable share. Third-party market research generally lists SCREEN as the global leader in wafer cleaning equipment: in single-wafer cleaning, the most important advanced-node subsegment, one study gives SCREEN about 31% share through its SS series and says SCREEN also leads in wet bench batch systems; another single-wafer cleaner report estimates SCREEN’s global revenue share at about 25%, ranking it first in the niche. Share estimates range from 25%–31%, but the qualitative “cleaning number one” is consistent across sources and supports the report’s cited Gartner 2024 view that SCREEN ranks number one globally in the three cleaning categories.

  The reason this moat cannot be copied by “buying a few machine tools” lies less in the hardware itself and more in the customer qualification chain: particle control, chemical-fluid management, accumulated yield data, stable operation on mass-production lines, and once obtained, PoR (plan-of-record) status. These require years of joint development with customers. The report’s statement that “the difficulty of replication lies more in process qualification time and customer trust than in simple capital spending” is sound. Pricing power also has hard evidence: even with FY2026 revenue down -3.1%, operating margin held at 20.2%, and SPE still generated ¥122.7bn of operating profit despite revenue falling -6.5%. That shows it is not a price taker in critical processes.

  But the boundary of the moat must be stated honestly: this is process-critical pricing power, not consumer-brand pricing power; it is a deep moat around a single process step, not a broad moat across multiple process categories. Its indispensability is concentrated in cleaning/surface treatment. Outside that step, it does not have TEL’s category depth across coating/developing/etch/deposition/cleaning, nor KLA’s inspection/metrology dominance. Deep, but narrow.

  ---

  2. Forces widening the moat: advanced processes/advanced packaging/HBM make the cleaning step more valuable

  This is the strongest part of the SCREEN bull case, and it is genuinely improving. Physically, below 7nm, single-atomic-layer contamination is enough to sharply increase device leakage and reduce reliability, posing a “severe challenge” for cleaning. Industrially, single-wafer cleaners already accounted for more than 35% of the cleaning market in 2025 because of their precision, uniformity, and suitability for advanced nodes, while the entire wafer cleaning equipment market grows from about $6.42 billion in 2025 to about $9.92 billion in 2031. Advanced packaging/chiplets/HBM introduce more hybrid bonding, TSV, thin-wafer, and related steps, each requiring cleaning, with cleanliness requirements only increasing.

  Implication for SCREEN: the cleaning step’s “process weight” and unit value in the whole line rise, customers become more willing to pay a premium for qualified process solutions, and SCREEN sits precisely in the leading position in the most critical single-wafer cleaning market. This is the real basis for the report’s “core cleaning moat has slightly widened” judgment. What widens is depth/unit value/stickiness, not breadth of market share. This distinction matters: a more valuable process step makes each order more profitable and customer switching harder, but it does not automatically expand the customer base or geographic scope SCREEN can serve.

  ---

  3. Forces narrowing the moat, part 1: Chinese local substitution. This is the most concrete and immediate threat

  The report is right to list China exposure of 42.4% as a core risk, and the threat is named and already happening, not abstract:

  • ACM Research: it started with cleaning as its core business and directly attacks SCREEN’s backyard. FY2025 revenue was about $782 million, up about 40% year on year, driven by installed-base expansion at Chinese fabs at 28nm and below; in March 2025, its high-temperature SPM tool was certified for sub-28nm nodes. It is already in the global competitive tier for cleaning equipment (SCREEN/TEL/AMAT/ACM/Lam together account for about 65% of 2024 cleaning revenue). More concerning is its long-term target: more than $4.0 billion revenue, with about $2.4 billion from mainland China and $1.6 billion from overseas. That means it does not only want to eat the Chinese market; it wants to export qualified processes overseas, including Singapore and Oregon in the US, and directly attack SCREEN’s international customers.
  • Naura: by 2024 it was already the world’s sixth-largest semiconductor equipment vendor. Cleaning, single-wafer plus bath systems, is one step in its process chain, and through acquiring Kingsemi it added coating plus cleaning, forming an etch-thin-film-cleaning full-flow solution. Its threat is not single-point cleaning share; it is domestic whole-line bundling. When Chinese fabs can source domestic etch + deposition + cleaning in one package, SCREEN’s single-point advantage in China is diluted by packaged substitution.

  Net effect: in China, which accounts for 42.4% of SCREEN revenue, the company faces a triple squeeze of subsidies, localization policy, and technology catch-up by local vendors. The catch-up has already reached 28nm, the node opponents care most about. The breadth boundary of this moat is being cut away in concrete terms, and the direction is irreversible. This is the main weight pulling the net effect back from “clearly widening” toward “neutral.”

  ---

  4. Forces narrowing the moat, part 2: dry-process substitution of wet processes. Real but gradual, not disruptive

  The company itself lists “wet cleaning/etch being replaced by dry processes” as a risk in its annual report SWOT, and this needs to be acknowledged honestly. Trend data also supports “dry is growing”: wet cleaning still led in 2025 with about 40% share, but dry cleaning is expected to grow faster because it offers higher precision at advanced nodes, can handle organics/nitrides/oxides, and reduces wastewater; dry processes are described as “necessary” for removing certain contaminants in logic/memory at 28nm and below.

  But the magnitude must be judged correctly: this is structural share shift, not disruptive replacement. Wet cleaning remains the largest part of cleaning, at ~40%. In advanced processes, dry and wet methods are more often complementary, each suited to different contaminants and different steps, rather than dry replacing wet wholesale. At the same time, single-wafer wet cleaning itself is expanding because of advanced nodes, with single-wafer share rising above 35%. So the real threat to SCREEN is long-term edge erosion plus the need for continuous R&D to preserve wet cleaning’s role in critical steps, not the core market being taken away overnight. Over 3–5 years, this is a slow variable, negative in direction and limited in magnitude, but it cannot be ignored. Combined with Chinese substitution, it caps the “breadth” of the moat.

  ---

  5. Compared with broad-category leaders (TEL/Lam/AMAT/KLA), SCREEN is deep and narrow

  Place SCREEN back into the competitive landscape and the positioning is clear:

  Dimension	SCREEN	TEL / Lam / AMAT / KLA
  Moat shape	Single-process deep moat (cleaning leader)	Cross-category broad moat (depth across multiple processes/equipment types)
  Indispensability	Concentrated in the cleaning step	Spans etch/deposition/coating-developing/metrology, with broader customer dependence
  Substitution resilience	Vulnerable if a single process route, wet process, is substituted	Multiple-product portfolios diversify single-process substitution risk
  Market valuation	trailing P/E ~28x, forward ~22x	TEL ~49x, ASML ~56x, KLA ~55x P/E

  The report’s comparison is accurate: SCREEN’s relative cheapness versus top-tier leaders partly reflects the market’s discount for “narrowness.” Broad-category leaders have stronger indispensability and more stable capital returns, so they enjoy higher multiples. SCREEN’s moat is real and its margins are strong, but it does not have the compounding expansion quality of becoming wider over time and automatically consuming adjacent process steps. Its growth depends more on a single process step becoming more valuable and retaining number-one share, not category spillover.

  ---

  6. Final net judgment, honest and without overstating the growth narrative

  Combine the four forces:

  • Widening, from advanced processes/packaging/HBM raising cleaning process weight and unit value, is real and acts on the moat’s depth;
  • Narrowing, from accelerated Chinese substitution plus gradual dry-process erosion, is real and approaching, and acts on the moat’s breadth.

  The two do not offset on the same dimension: depth is increasing, breadth is being cut. My net judgment is that the core-process moat slightly deepens, the overall serviceable boundary slightly narrows, and the combined direction is near-neutral with a slight positive tilt only for the high-end share it can defend. The magnitude is not large, and the main uncertainty rests on the China variable. This aligns with the report’s conclusion that the moat is “overall stable, with core cleaning moat slightly widening and non-core businesses weaker,” but I would add a more restrained sentence: this moat will not “automatically widen over time until competitors cannot catch up.” SCREEN must actively maintain it through R&D, advanced-node qualification, and overseas customer expansion. If it loses share in China or falls behind on dry-process routes, the depth gains will quickly be consumed by breadth losses.

  So, returning to the literal question, “Will the moat widen or narrow over the next three to five years?” Core high-end cleaning process: slightly wider, meaning deeper. Overall market coverage and single-process safety: pressured and narrower. For a stock rated “Watch: good company, but not at a good enough price,” this is the key: the moat is real enough to support “good company,” but its “deep and narrow, with pressure from both sides” nature means it does not deserve the valuation premium of a broad-category leader. That echoes the report’s core stance: “relatively cheap does not mean absolutely cheap, and the current price has already overdrawn the cycle.”

  Jun 5, 2026
• QIf its core business is disrupted, does it have the DNA to reinvent itself? How does it handle mistakes and bad news?4/10

  A

  Bottom line: the reinvention DNA is real across the long arc of history, but limited in the current dry-process substitution context. Over more than eighty years, SCREEN migrated across industries from screen-making photographic equipment to metal mesh for television camera tubes to semiconductor lithography/cleaning equipment, proving that the company has a genuine historical record of surviving by changing fields. This is not empty talk. But directly applying that DNA to “wet processes being replaced by dry processes” as an upside narrative is not well supported by evidence. Its response looks more like deepening the wet-process/cleaning moat and collaborating with dry-process leaders, rather than growing its own second dry-process leg. On bad news, its disclosure culture is clearly above industry average: it proactively lists inventory valuation as a key audit matter, writes adverse risks such as “share loss” and “wet processes being replaced by dry processes” into its annual-report SWOT, and when FY2026 turned down it directly attributed the result to “higher fixed costs + lower revenue” rather than dressing it up. Overall: reinvention DNA is historically credible but not yet proven today; bad-news handling is honest and earns credit. Still, honestly, neither point solves the core conflict of “strong cyclicality + valuation already carrying a premium.” Under the Baillie Gifford framework, this is not the kind of stock where the market has failed to understand the years 3–10 path to a 5 times outcome.

  ---

  ① Reinvention DNA: the century-long evolution is real; the dry-process hedge is weak

  The historical evolution stands up to verification and is richer than the report’s body suggests. The company’s roots go back to Ishida Kyokuzan Printing Works in 1868. In 1934, it produced Japan’s first glass screen for photographic plate making. In 1943, it began operating independently as Dainippon Screen, initially as an integrated photographic plate-making/screening equipment manufacturer. In 1955, it migrated its fine-line screen engraving technology into metal mesh for television camera tubes, entering electronic components. In the mid-1970s, it launched its first semiconductor lithography system, formally entering semiconductor equipment. It changed its name to SCREEN Holdings only in 2014. This is a real cross-technology evolution chain: photographic plate making to electronic components to semiconductor equipment. Core capabilities, precision patterning, surface treatment, and microfabrication know-how, were reused while the carrier industries changed several times. Baillie Gifford’s prized “self-reinvention DNA” is valid on this century-long timeline. It at least proves this is not a one-trick tool vendor; it has a successful precedent of leaving its original industry and using underlying process capabilities to build a new main business.

  But “reinvented itself historically” does not mean “can reinvent itself this time.” The Baillie Gifford framework needs forward-looking response evidence, and SCREEN’s current hedge against the specific threat of “wet processes being replaced by dry processes” looks limited and defensive to me:

  • The most concrete new evidence is that on May 26, 2026, Applied Materials announced SCREEN had joined its EPIC Center, combining SCREEN’s single-wafer cleaning, wet etch, and surface preparation with Applied’s deposition, etch including dry processes, and materials modification for co-optimization at the most advanced nodes. SCREEN executive Akihiko Okamoto’s own words were that “the interfaces between wet etching/cleaning and adjacent process steps have never been more important.” That sentence itself acknowledges that at advanced nodes, wet processes can no longer close the loop independently and must work tightly with dry processes.
  • The essence of this move should be seen clearly: it deepens SCREEN’s wet-process/cleaning moat and embeds it into advanced-node process integration, while binding it to Applied as a dry-process leader. It does not mean SCREEN has grown its own dry-etch capability. In other words, this is a defensive action of reinforcing the core and allying with the disruptor, not incubating a second curve that can replace the core business.
  • The industry trend is not friendly, but it is not that bleak either: dry/low-temperature (cryogenic)/supercritical CO₂ and other low-water or waterless processes are indeed penetrating faster in 3D NAND and other advanced structures, and rivals such as ACM Research are also working in these niches. This is exactly why the report and the company’s SWOT both name substitution risk. But balance is needed: cleaning/surface treatment is becoming more important in visible advanced logic, HBM, and advanced packaging; wet processes still have steps where dry processes struggle to fully replace them, especially in particle control and chemical selectivity; SCREEN’s installed base of more than 15,000 shipped cleaning tools and process-qualification barriers mean “disruption” is a slow variable, not a cliff.

  Summary of the first half of the question: reinvention DNA is real and credible historically, but unproven in the current context. Baillie Gifford would appreciate the century-long migration, but would not use it to excuse “dry-process substitution,” because the evidence available today points to reinforcing wet processes plus partnering with dry-process leaders, not self-incubating dry processes. The evidence lacks a clear “next active jump to a new field.” Honest score: neutral to weak, not an upside catalyst.

  ② How it handles mistakes and bad news: honest, and this is a real positive

  On this dimension, SCREEN performs clearly better than the average “strong-cycle equipment maker,” and the evidence is solid without embellishment:

  • It proactively exposes the most sensitive accounting weak point. The auditor lists the assessment of salability for SPE-related finished goods and work-in-process as a key audit matter. Inventory valuation is exactly where trouble most easily appears when the cycle turns. The company did not avoid it; it placed it where investors can see it clearly. This is a typical act of putting bad news on the table.
  • It writes adverse risks against itself. In its annual-report SWOT, the company lists “loss of share from competitors’ technology and pricing improvements,” “wet cleaning/etch being replaced by dry processes,” and “geopolitics” as risks. In other words, the dry-process substitution threat tested in this question was acknowledged first by the company itself, not imposed from outside. A company willing to list “the moat may be filled in” against itself has relatively high governance integrity.
  • It does not dress up a down year. In FY2026, revenue fell -3.1%, operating profit fell -9.7%, and net income fell -7.5%. Management directly attributed this to the simple and falsifiable operating explanation of “higher fixed costs + lower revenue,” rather than resorting to one-off, vague excuses. Dividends also honestly declined from ¥308 in FY2025 to ¥293 in FY2026, with the market forecasting a further decline toward about ¥175 in FY2027, recognizing a new phase of the cycle instead of forcing a “linear growth” persona.

  This culture of “stating bad news before the share price and before external critics force it out” is exactly the kind of quality Baillie Gifford values when judging whether management can accompany shareholders over the long term. Honest score: handling mistakes and bad news is candid and positive.

  ---

  Overall framing for question 5, without inflating the growth narrative: SCREEN has both “it really reinvented itself historically” and “it is genuinely honest about bad news” as underlying traits, which makes it qualified and even somewhat above average as a long-term steward. But the core of Baillie Gifford’s framework is upside imagination for a fivefold result over the next ten years, and the real upside variable in this question, whether SCREEN can reinvent itself again in the face of dry-process substitution, currently has only defensive positioning and no offensive proof. Add the report’s main Buffett-style framework of “Watch: good company, but not at a good enough price,” strong cyclicality already visible in FY2026, and a valuation that already incorporates AI strength and FY2027 recovery, and the honest conclusion is that this company’s reinvention DNA and honest culture are real protective qualities, but not the engine for a 5 times upside. It is more like a high-quality cyclical leader worth respecting and tracking at a good price, not a great growth stock that the market has failed to understand in the Baillie Gifford sense.

  Jun 5, 2026
• QDoes management, especially the founder, have a long-term view and deep alignment with the company? Is it willing to sacrifice current profit for five to ten years from now?4/10

  A

  Bottom line: on the Baillie Gifford dimension of “management long-term view + deep alignment,” SCREEN is a case of good governance discipline but weak depth of alignment. It is a century-old company founded in 1943 and now fully professional-manager-run. It is clearly not founder-led or owner-operated. The personal shareholdings of the three core executives amount to far below 0.01% of company equity, making it clearly weaker than the Baillie Gifford template of a founder staking personal wealth and willingly sacrificing current profit for ten years out. But it does well on the other leg: measurable capital-allocation discipline and shareholder-oriented governance, with compensation linked to ROIC, net cash, buybacks and dividends, and explicit medium-term return targets. This is “institutionalized restraint,” not a founder-led expedition. For an LTGG framework, this is “a good steward you can trust, but not a captain betting ten years for you.”

  1. Clearly not founder/owner-operator, with very shallow alignment. SCREEN was founded in 1943 and is a typical century-old professional-manager enterprise. In June 2025, it completed a CEO transition, with Masato Goto becoming President and CEO. He joined in September 1990 and was promoted after more than thirty years in the semiconductor equipment business, making him a pure internal professional manager, not a founder or major shareholder in any sense. In terms of shareholding depth, the figures cited in the report from the general meeting notice, Chairman Hiroe 53,468 shares, CEO Goto 55,972 shares, CFO Kondo 19,816 shares, are already small. At the current price of about ¥12,605, Goto’s stake is worth only about ¥700 million, representing about 0.003% of the company’s equity relative to a market value of about ¥2.54 trillion. The three together are far below 0.01% of company equity, with supporting evidence that former CEO Hiroe’s direct stake was once only about 0.028%. That means executives’ personal wealth is only very lightly tied to the stock price. This is exactly where Baillie Gifford would deduct points: it lacks the ownership structure where a founder has most of their wealth in the company and naturally thinks in ten-year terms. Institutions, such as Nomura Asset Management at about 7.5% as the largest shareholder and institutions collectively at about 53%, are the true ownership center. This is a widely held company run by professional managers as fiduciaries.

  2. Governance quality is a positive, but it is institutional rather than founder-like. Of 8 directors, 4 are independent outside directors, or half the board, and the company has nomination and compensation advisory committees dominated by outside directors. Compensation consists of fixed pay, short-term performance cash bonuses, and stock-based compensation linked to short-/medium-term performance and corporate value. The company explicitly links director incentives to ROIC, operating margin, and environmental/safety indicators. ISS Governance QualityScore was 3 in February 2026, with compensation and audit pillars both at the best score of 1. For a Japanese equipment company, this framework is quite advanced. Baillie Gifford would appreciate incentives aligned with long-term capital returns, but would also note the flaw cited in the report: final authority over individual director compensation is delegated to the CEO, constrained only by advisory committee input, which is less pure than a fully independent committee-led process. Also, ROIC and margin metrics lean toward capital discipline rather than the kind of long-termism Baillie Gifford most wants to see: sacrificing current reported profit for disruptive growth in years 3–10.

  3. Capital-allocation discipline: strong, but oriented toward steady returns rather than sacrificing today’s profit for the long term. This is one of SCREEN’s strongest recent points. The medium-term plan Value Up Further 2026 (FY2025–2027) explicitly targets average operating margin above 19%, ROIC above 15%, dividend payout ratio above 30%, and flexible buybacks, while positioning the three years as an “investment period” serving a longer-term goal of ¥1tn sales and 20% operating margin ten years later. In execution, FY2025 ROIC reached 24.7%, FY2026 total shareholder return was 42.3%, the company repurchased 1,242,500 shares during the year for ¥11.1bn, and it maintained a net-cash structure, with cash of about ¥225.7bn versus interest-bearing debt of about ¥4.6bn, operating almost entirely with internal funds. This is rational “good steward” behavior: it does not chase scale with leverage, and it combines buybacks and dividends during high-profit periods. But two points are unfavorable from a Baillie Gifford perspective. First, it reflects restraint: returning cash to shareholders and protecting the balance sheet at a cyclical high, rather than depressing current profit and reinvesting cash into a new curve capable of supporting a fivefold outcome over ten years. Second, the company lists R&D facility expansion as the main capex item, while profit has fallen with the cycle, with FY2026 revenue down -3.1% and net income down -7.5%; third-party figures show trailing ROIC falling from about 22.6% to about 18.7%. This shows its excellent returns have a meaningful cyclical component. Management has not, and does not appear inclined to, voluntarily sacrifice current profit for a bet on the future. That is the opposite of the Baillie Gifford template.

  4. Comparison with Baillie Gifford’s preferred profile. Baillie Gifford’s ideal profile is founder/owner-operator, deep shareholding, a view longer than ten years, and a willingness to sacrifice current profit for years 3–10 upside. Against that ruler, SCREEN scores as follows: long-term view: neutral, with a clear ten-year ¥1tn blueprint, but the narrative is gradual share expansion and steady returns, not a disruptive expedition; alignment: weak, with no founder and negligible executive ownership, so personal wealth is lightly tied to the share price; willingness to sacrifice current profit for the long term: weak to neutral, because ROIC/margin/payout ratio are hard constraints and management tends to preserve current returns rather than depress profit for growth; governance quality and capital discipline: strong, with majority independent directors, ROIC-linked pay, net cash, and disciplined buybacks/dividends. Overall, this is a company with credible management and mature capital allocation, but it is weak on Baillie Gifford’s most prized features: founder deep alignment and long-termist boldness. It is more like a well-run fiduciary machine than an owner-operator with strong personal will leading the company through ten years of uncertainty for fivefold upside. For a growth investor seeking “fivefold over the next ten years,” this is not a positive; it is a structural limitation to accept clearly. You are buying governance and discipline, not founder-like foresight and boldness.

  Jun 5, 2026
• QIf it disappeared tomorrow, how much would customers miss it? Is its growth model sustainable and not dependent on harming society or regulation?4/10

  A

  Bottom line: customers would miss SCREEN a lot, but it is far from irreplaceable. If it disappeared tomorrow, advanced fabs’ cleaning lines would go through painful requalification and yield volatility, but TEL, Lam, AMAT, Korea’s SEMES, and China’s ACM could take orders. Customers already generally use dual or multiple suppliers, so SCREEN is a critical supplier that is painful but replaceable, not an ASML-like monopoly in EUV lithography or a Lasertec-like near-monopoly in EUV mask inspection. The growth model itself is healthy: semiconductor equipment does not harm society, and regulation generally helps by treating advanced processes as strategic assets. The real factor eroding its “sustainability” is not bad behavior by SCREEN, but the fact that 42.4% of revenue sits in China, which is the eye of the dual storm of export controls and local substitution, and cleaning equipment has been explicitly included in the US BIS control list. Overall, indispensability is upper-medium, not top-tier; social/regulatory sustainability of growth is structurally questionable, one of the core reasons this good company is held down at Watch. This question is weak.

  ---

  ① Indispensability: highly important, but not exclusive

  Cleaning is a real pain point. Every added chip structure requires repeated cleaning to remove particles, metals, and organic residues. Impurities directly hurt yield. As processes move further into advanced nodes, HBM, advanced packaging, and chiplets, cleaning steps increase and requirements become stricter. The report’s judgment that the moat has slightly widened is sound in industrial logic. SCREEN is indeed the global leader in this niche: third-party estimates put its single-wafer cleaning share in a range from about 42% to about 45% under ICS 2024 estimates, and it holds leading positions across batch cleaning and spin scrubbers as well. This externally supports the report’s “global number one in three cleaning categories + PoR status + high switching costs” view. Switching costs are real: changing suppliers requires repeating process introduction, chemical-fluid and particle-control qualification, and mass-production stability certification, which can take months and carries real yield risk. Once a customer installs SCREEN tools, it will not casually move away.

  But “being missed” has a ceiling because SCREEN is not exclusive. Wet cleaning is a structurally multi-supplier market: SCREEN, Tokyo Electron, Lam Research, and Applied Materials together hold more than 90% of revenue. If the scope is expanded to include Korea’s SEMES and China’s ACM Research, the top five account for about 65% of revenue. This means major customers such as TSMC, Samsung, and Micron already spread cleaning equipment across several vendors, and most critical steps have a second or third qualified supplier as a backstop. So the real consequence of “disappearing tomorrow” would be customers being forced to shift orders to TEL/Lam and other qualified or quickly qualifiable alternatives, absorbing 6–18 months of migration pain and short-term yield disturbance, not production lines shutting down.

  The gap versus the two benchmarks named in the brief is clear:

  • ASML: sole global EUV lithography-machine supplier. No substitute exists. If it disappeared tomorrow, global advanced logic/HBM manufacturing would directly stop. That is true chokepoint indispensability.
  • Lasertec: near-exclusive in EUV mask defect inspection (actinic). Advanced-process photomask qualification cannot bypass it.

  SCREEN is not in that league. It is the supplier with the highest share and most painful switching cost within a niche, but customers always hold TEL/Lam as backup cards. The report’s framing is accurate: “an excellent niche leader in a good industry, but not an invincible monopolist” and “process-critical pricing power rather than consumer-brand pricing power.” This also explains why its valuation multiple, trailing P/E about 27x, is significantly below ASML at about 56x and KLA at about 55x. The market does not give it an exclusive-monopoly premium because it is not exclusive.

  Indispensability summary: upper-medium. Strengths are painful switching, number-one share, and advanced processes raising cleaning weight. Weaknesses are customer dual/multiple sourcing, TEL/Lam/AMAT able to step in, and it is far from ASML/Lasertec-style exclusivity.

  ② Social and regulatory sustainability of growth: clean business, but exposure sits in the storm’s eye

  Start with the positive: the growth model does not depend on harming society. Semiconductor cleaning equipment makes chips more reliable and yields higher. It is an enabling layer for positive technology waves such as AI, HBM, and advanced packaging. It does not involve addiction, pollution arbitrage, regulatory arbitrage, or business models that profit by harming others. From the Baillie Gifford question of “is the money clean, and does it rely on harm,” SCREEN passes. Regulation generally encourages rather than suppresses semiconductor manufacturing because advanced processes are strategic assets for many countries. This is a real positive.

  But “sustainability” stumbles on the other side of geopolitics and regulation. Baillie Gifford asks whether growth is sustainable and not dependent on regulation. SCREEN’s growth is highly exposed to a form of regulation it cannot control:

  • China exposure is 42.4% in FY2025, Taiwan 18.1%, and Asia-Pacific about 69.5%, while China is the core target of global export controls. This is not “regulation caused by SCREEN doing something wrong”; it is “its largest single market is directly circumscribed by third-country regulation.”
  • Cleaning equipment has been explicitly included in controls. The latest US BIS rule lists etch, deposition, lithography, annealing, metrology/inspection, and cleaning tools together as controlled advanced-node manufacturing equipment. “Wafer manufacturing cleaning and removal equipment” is a named controlled item. Cleaning is not outside the control perimeter. At the same time, in August 2025, the US closed a loophole that had previously allowed some foreign-owned manufacturers to export equipment to China without licenses. As a Japanese vendor, SCREEN’s situation is now more closely aligned with US peers, and flexibility in supplying advanced nodes in China has narrowed.
  • Local substitution is the structural offset: China accounts for about 40% of global fab-equipment purchases, and local equipment share keeps rising. China’s ACM Research is itself a direct cleaning-equipment competitor. In other words, SCREEN faces a squeeze in China from both sides: regulatory ceilings above and local-substitution floors below. This is exactly why the report lists “accelerating Chinese local substitution + regulation eroding core market position” as a permanent-capital-loss scenario and includes it in sell/reassessment triggers.

  Two points need to be separated honestly: first, exports of mature-node cleaning equipment to China can still continue for now, and tighter controls mainly hit advanced nodes, so this is a structural risk that gradually adds pressure, not an overnight cliff; second, this risk is common across Japanese and US equipment chains, not a SCREEN-specific flaw. But the concentration of 42.4% makes SCREEN’s sensitivity to the shock larger than peers.

  Sustainability summary: the business itself is clean, and regulation broadly supports semiconductor manufacturing, which is positive; but growth is heavily tied to a market defined by third-country export controls and facing local substitution, while cleaning equipment is already on the controlled list. This is a structural sustainability constraint that SCREEN cannot remove through its own operations.

  ---

  Overall judgment for question 7: weak. If SCREEN disappeared tomorrow, customers would suffer, but they could switch. It is critical, not a lifeline, and does not deserve an ASML/Lasertec-style indispensability premium. On whether growth is sustainable and not dependent on harming society or regulation, it passes the “not harming society” test, but is badly dragged down by “nearly 70% of revenue in the most geopolitically sensitive region, cleaning equipment already controlled, and local substitution advancing step by step.” This is one micro-level explanation for the report’s “good company but not at a good enough price” view at the current price of about ¥12,605 and market value of about ¥2.54 trillion: you pay a high price for its cleaning-leader position, but you do not buy an exclusive monopoly or a growth source insulated from geopolitical regulation.

  Jun 5, 2026
• QWhat are the unit economics of this business, including gross margin and incremental returns? Does scale make the business better or worse? Where does the money it earns go?5/10

  A

  Bottom line: this is a business with excellent absolute economics, but not top-tier relative to peers, and its profit elasticity includes a cyclical tailwind. In manufacturing, the unit economics are first-class: FY2026 gross margin of 38.5%, operating margin of 20.2%, FY2025 ROIC of 24.7%, and textbook positive operating leverage over the past five years, with revenue CAGR around 13.6% but operating profit CAGR around 38% and net income CAGR around 43%. Share gains and scale absorption have indeed structurally lifted margins. But three things need to be admitted honestly: first, its gross margin and operating margin are only mid-to-upper in global equipment leaders and clearly below KLA and Lam, while roughly in line with or slightly below AMAT and Tokyo Electron; second, reverse leverage already appeared from FY2025 to FY2026, with revenue down -3.1% but operating profit down -9.7% and net income down -7.5%, proving that these high margins amplify both upside and downside, and include a meaningful cyclical peak benefit from AI/HBM/advanced packaging plus Chinese buying, not pure structural monopoly; third, where the money goes is rational but not optimal: disciplined dividends and buybacks, net cash to resist cycles, but more than ¥225.0 billion of net cash sitting on the balance sheet while the highest incremental return comes from reinvesting in the business itself, with ROIC of 24.7% far above the opportunity cost of holding cash. This makes capital allocation “safe but not aggressive enough.” For growth investors, the unit economics are good enough to warrant long-term tracking, but its margin ceiling and cyclical beta mean it is not a compounding machine where scale automatically thickens the moat and margins rise monotonically.

  Four layers below.

  1. Absolute level: a standout in manufacturing, middle of the pack among equipment peers

  SCREEN’s unit economics are excellent in any “physical manufacturing” coordinate system: FY2026 gross margin 38.5%, operating margin 20.2%, net margin attributable to owners about 15.2%, FY2025 operating margin even higher at 21.7%, ROIC 24.7%, and ROE 25.1%. A company selling physical cleaning equipment to fabs and still generating 20%+ operating margin and 24.7% ROIC is capturing not assembly value, but the value of process know-how and PoR (process of record) status. The report argues this well. Even with FY2026 SPE revenue down 6.5%, operating profit was still ¥122.7bn, so pricing power is real.

  But in the narrower semiconductor-equipment coordinate system, it is only mid-to-upper, not elite. Compare the latest available peer figures, using each company’s most recent fiscal year or trailing periods:

  • KLA: gross margin about 62%, operating margin about 43%, and ROIC about 38%. Inspection/metrology oligopoly economics dominate the field.
  • Lam Research: recent-quarter GAAP gross margin about 49.8%, operating margin about 35%, as an etch/deposition leader.
  • Applied Materials: FY2025 gross margin 48.7%, operating margin 29.2%, and ROIC about 28%.
  • Tokyo Electron: FY2026, ending 2026/3, gross margin about 45.3%, operating margin about 25.6%, ROE near 30%, and ROIC about 22%–24%.

  The readout is clear: SCREEN’s gross margin of 38.5% is the lowest tier in this group, and its operating margin of 20.2% is below all four peers above. This reflects category structure. KLA sells high-value inspection/metrology with near-monopoly characteristics; Lam/AMAT are multi-category platforms with broader bargaining surfaces; SCREEN concentrates its firepower on the cleaning step, and the step’s substitutability risk, wet to dry, and value share within the process flow set a lower gross-margin ceiling than inspection and etch. The one metric where it enters the upper tier is ROIC at 24.7%, helped by light capex, only ¥27.7bn in FY2026, net cash, and operating efficiency. ROIC is roughly in line with TEL/AMAT and clearly above the industry average, trailing only KLA. So the honest picture is: mid-level margins, upper-tier capital returns, excellent absolute economics, but not “the world’s strongest unit economics.”

  2. Operating leverage direction: upside leverage has been proven, but reverse leverage is equally real. This is the key honest point

  On the question “does scale make margins better or worse,” SCREEN gives a two-way, symmetric answer, and growth investors must see both sides:

  • Upside, FY2021 to FY2025/26: operating margin climbed from 7.6% to 21.7% and then 20.2%; five-year revenue CAGR of about 13.6% drove operating profit CAGR of about 38% and net income CAGR of about 43%. Fixed-cost and R&D absorption, share gains, and product-mix improvement translated scale expansion into a margin jump. This is real structural improvement and a genuine part of moat monetization.
  • Downside, FY2025 to FY2026: revenue fell only -3.1%, but operating profit fell -9.7% and net income fell -7.5%. Management directly attributed it to “higher fixed costs + lower revenue.” This is the first clear appearance of reverse operating leverage. The same high-fixed-cost structure that helps on the way up bites immediately when revenue turns down.

  This means the business’s high margin is not a compounding curve where scale makes everything monotonically better; it is a curve that swings with the capex cycle, and leverage amplifies the swing. From Baillie Gifford’s years 3–10 perspective, this is a negative: you cannot assume FY2025’s 21.7% operating margin is a baseline that can be linearly extrapolated. A more realistic midpoint may fluctuate around 17%–20% through the cycle. Management’s own Value Up Further 2026 medium-term plan targets average operating margin above 19% and ROIC above 15%, clearly below FY2025’s peak, which amounts to official acknowledgment that the peak is not sustainable.

  3. Within high margins: how much is structural, how much cyclical tailwind?

  This is where the question needs a plain answer. My judgment is seven parts structural, three parts cyclical, but the three cyclical parts are enough to distort extrapolation:

  • The structural part, credible: global number-one share across the three cleaning segments, single-wafer/batch/spin scrubber, PoR status, process-qualification barriers, and advanced packaging/HBM raising the importance of cleaning. These support a margin baseline above ordinary equipment makers. Holding a 20% operating margin even in a down year proves the baseline is real.
  • The cyclical tailwind, needs a discount: from FY2024 to FY2025, revenue surged from ¥504.9bn to ¥625.3bn (+24%), and operating margin jumped from 18.6% to 21.7%. That jump included dual tailwinds from AI/HBM expansion and large-scale Chinese buying, with China at 42.4% of FY2025 revenue. The report’s bear case says it well: “You are buying a good company, but at the intersection of a strong cycle and high expectations.” If AI capex slows and Chinese local substitution scales, as Chinese domestic equipment share keeps rising, this tailwind can retreat. A margin midpoint of 17%–19% and ROIC returning toward 15%–20% are reasonable scenarios. FY2026’s turn, with operating profit down -9.7%, is a preview.

  Conclusion: treating FY2025’s 21.7% operating margin and 24.7% ROIC as the structural norm would overstate this business. Using management’s own 19% operating margin / 15% ROIC as the midpoint and treating the peak as a cyclical bonus is the honest growth assumption.

  4. Where the money goes: rational, but net-cash hoarding may not be optimal

  Start with FCF conversion: FY2026 operating cash flow was ¥92.7bn and free cash flow was ¥62.9bn, while average FCF/net income over the past four years was about 75%. This conversion shows accounting profit largely turns into cash, but this is not a light-asset model where profit equals cash. Inventory, about ¥168.7bn in FY2025, and contract-liability fluctuations can move single-year cash flow up and down, and the salability of SPE finished goods/work-in-process is a key audit matter. When the cycle turns, inventory write-down sensitivity matters. So FCF quality is good but not top-tier.

  The money has four uses, assessed one by one:

  1. Dividends: ¥308 in FY2025 and ¥293 in FY2026, with disciplined payout ratio above 30%. Note, however, that under Yahoo Japan figures, the FY2027 dividend forecast falls to ¥175, meaning management uses a dividend cut to acknowledge the cycle entering a new phase. This is an honesty signal, not capital-allocation deterioration.
  2. Buybacks: ¥11.1bn in FY2026 and ¥30.0bn over the 2025 round, buying back stock during a high-profit period while maintaining net cash. Rational.
  3. Reinvestment, with capex of ¥27.7bn mainly into SPE R&D facility expansion: this is the highest-incremental-return use and the one that deserves more capital. ROIC of 24.7% is far above the opportunity cost of any other funding use. Putting more money back into a business with 24.7% ROIC is the path to maximizing compounding. The report is right that a meaningful share of this capex is growth capex.
  4. Net-cash accumulation, with cash of ¥225.7bn versus interest-bearing debt of ¥4.6bn, plus a retained ¥60.0bn credit line: this is the most debatable item. For a cyclical stock, a thick net-cash position has real downside value: it helps survive cycles, avoid forced selling, and avoid debt-funded buybacks. FY2026 operations funded entirely by internal resources prove this. But from a growth/compounding-efficiency perspective, ¥225.0 billion of cash earning roughly a little above 2.6% on Japanese government bonds creates a huge opportunity-cost gap versus 24.7% ROIC in the core business. A truly aggressive compounding machine would either reinvest more in the high-ROIC core, buy back more aggressively, or conduct high-return acquisitions within its circle of competence. Keeping such a large share of capital idle in 2% assets for a long time is a “safe but capital-inefficient” compromise. The opposing argument also holds: under strong cyclicality, China geopolitics, and dry-process substitution uncertainty, retaining ample ammunition is a reasonable price for avoiding permanent loss. So this item is a balance of “defensively correct, offensively conservative”; not optimal capital allocation, but not wrong either.

  One-sentence close for the growth perspective: SCREEN’s unit economics are strong enough to put it on a high-quality watch list: upper-tier ROIC, proven upside operating leverage, solid FCF conversion, disciplined capital allocation, and net cash to resist cycles. But its gross margin/operating margin are only middle-of-the-pack among equipment peers, profit elasticity includes a cyclical tailwind that can retreat, and the largest pool of capital sits in 2% cash. This is a very good but not extreme business, one that breathes with the cycle rather than moving monotonically upward. That is exactly what limits its eligibility as Baillie Gifford’s “fivefold in ten years, margins widening monotonically” great compounding machine.

  Jun 5, 2026
• QWhat conditions must hold simultaneously for it to rise fivefold over ten years? Are those conditions realistic? What expectations are embedded in today’s share price?3/10

  A

  Bottom line: through the Baillie Gifford LTGG lens, the probability of SCREEN (7735.TSE) rising fivefold over ten years is low. Moving from a current price of about ¥12,605 to about ¥63,000 requires about 17.5% annualized. That requires both the “earnings multiple” and the “valuation multiple” wheels to work. But SCREEN’s current facts are: earnings have just turned down in FY2026, with revenue -3.1% and net income -7.5%; five-year revenue CAGR is only 13.6% and the operating-leverage release phase has passed; and the starting valuation, with forward P/E already 22.18x, is high rather than low. Either the already-not-cheap valuation must be pushed much higher, or earnings growth over the next ten years must be even faster than the past five years, which included an explosive share-gain phase. Both conflict with Baillie Gifford’s target of “nonlinear upside the market has not yet realized.” This does not mean SCREEN is not a good company; it means today’s price has prepaid many good things, leaving little fuel for a fivefold result.

  ---

  1. Break “fivefold” into two conditions that must both hold

  Ten-year fivefold = earnings growth multiple × valuation multiple change. Starting from current forward P/E ≈ 22.2x, several internally consistent combinations are possible (sensitivity illustration, not precise forecast):

  Path	Required ten-year EPS growth	Implied EPS CAGR	Required terminal P/E	Realism
  A: valuation unchanged, all earnings	About 5.0x	About 17.5%	Still ≈ 22x	Earnings requirement is extremely high
  B: earnings 3x + valuation expansion	About 3.0x	About 11.6%	About 37x	Double-high, and the cycle peak must not compress valuation
  C: earnings 2.5x + valuation expansion	About 2.5x	About 9.6%	About 44x	Valuation must reach the current Tokyo Electron 49x / ASML 56x tier

  Note: path C’s “44x” is compared with the report’s Tokyo Electron about 49x and ASML about 56x P/E range. In other words, for SCREEN to achieve a fivefold ten-year result via the “lower earnings + valuation re-rating” path, the market must be willing ten years from now to price it as a top-tier, most irreplaceable equipment leader.

  The table’s meaning is direct: if one is unwilling to assume the valuation multiple doubles, earnings must grow close to 5 times, or about ~17.5% annualized; if one is unwilling to assume earnings grow close to 5 times, one must assume a major valuation multiple expansion. Baillie Gifford does not reject valuation expansion, but its true bet is on nonlinear earnings explosion supporting valuation, not on market sentiment alone pushing the multiple from 22x to 40x+.

  2. Are these conditions realistic versus SCREEN’s facts? A line-by-line check

  • The earnings wheel, the most important one, is weaker than what “fivefold” requires. Past five-year revenue CAGR of 13.6% and net income CAGR of about 43% look very strong, but the report honestly notes that this was the release of three forces: good industry + strong cycle + share improvement. The operating margin moving from 7.6% in FY2021 to 20%+ is operating leverage that can only be used once. Margins are already high, leaving little room for another “margin doubling” to contribute growth. More realistically, FY2026 revenue is already -3.1% and net income -7.5%, with FY2026 revenue of ¥605.75bn down -3.12% year on year confirmed by actual data. For EPS to achieve ~17.5% annualized over ten years under path A, the company would need to be faster over the next ten years than over the past five years, which included one-off tailwinds, and without another meaningful down cycle. For an equipment business the report repeatedly stresses cannot escape cycles, this is hard to assign a high probability.

  • The market-share ceiling limits the earnings multiple. SCREEN’s strengths are cleaning/surface-treatment niches, where it is global number one across single-wafer, batch, and spin scrubber cleaning. But it is already global number one in these niches. The Baillie Gifford-favored share leap from 5% penetration to 30% does not have much room in its strongest markets. It is more likely to “hold number one + grow with industry beta” than to “take share and recreate another SCREEN.” The second curve, advanced packaging cleaning/display/printing, has imagination, but scale and certainty are insufficient to carry a “fivefold” result independently.

  • Buybacks can modestly amplify per-share metrics, but cannot fill the gap. Net cash and disciplined buybacks, such as the FY2026 ¥11.1bn repurchase, can make EPS grow slightly faster than net income. But this amplification is tiny versus the gap created by “earnings must go 5x,” and buyback economics worsen at high valuations.

  • The valuation wheel is currently a headwind, not a tailwind. Starting forward P/E of 22.2x, trailing P/E 27.6x, P/B 5.1x, and P/FCF 39.6x are not “undervalued waiting for re-rating” levels. Sell-side consensus 12-month target price is about ¥12,790, almost flat to and even slightly below the current price, while third-party long-term growth expectations are only earnings about +2.1%/year and revenue about +3%/year, slower than Japan’s market at 8%/year. In other words, external consensus itself does not believe in the high growth needed for “fivefold.” That means achieving fivefold requires the valuation multiple not only to avoid compression, but to expand against today’s low growth expectations, making the challenge greater.

  3. What does today’s ¥12,605 share price imply?

  • It does not imply “fivefold,” but it already implies the relatively optimistic assumption of “cycle continuation + FY2027 recovery delivery.” The report’s neutral DCF intrinsic value is only ¥8,300, and the optimistic case only ¥12,200, while the current price is already close to or even slightly above the upper edge of the optimistic value. The report estimates ten-year annualized returns from buying at the current price as conservative -3%~0%, neutral 3%~5%, optimistic 7%~10%. Note: even the report’s optimistic scenario annualized return of 7%~10% does not reach the 17.5% needed for a fivefold result. In other words, at the current price, the market has already incorporated good news such as a cycle upswing, stable share, and sustained AI capex into a level where a conservative investor has almost no margin of safety. It implies a sentiment premium for a “high-quality growth equipment stock,” not an unrecognized asymmetric fivefold opportunity the market has missed.

  • From Baillie Gifford’s reverse question of “why has the market not realized this,” there is no obvious cognitive gap here. Sell-side coverage is sufficient, with 15 analysts, a Buy rating, but a nearly flat target price. The AI/HBM/advanced packaging narrative is already consensus and priced in, and the 6/3 single-day +17.94% surge shows sentiment is already quite excited. Baillie Gifford wants a view that is correct but unpopular; SCREEN today looks more like “correct and already popular,” which is the worst starting position for “fivefold.”

  4. Honest assessment: low probability

  Put the two wheels together: achieving close to 5x earnings under path A conflicts with the three facts that margins are already high, share is already number one, and the cycle has already turned down. Expanding valuation to 37–44x under paths B/C requires the market, against current low growth consensus, to re-rate a strongly cyclical niche leader into the tier of the top monopolistic equipment leaders. Each path requires a sequence of optimistic conditions to hold simultaneously, and they are hard to combine: if earnings truly explode, it likely occurs near a cycle peak, when valuation is more likely to compress than expand. Therefore, through the Baillie Gifford lens, SCREEN’s probability of rising fivefold over ten years is low. It is more likely to be an investment with mid-single-digit annualized returns, high in quality but hurt by entry price, rather than the “fivefold in ten years, not yet priced by the market” great growth stock Baillie Gifford seeks. The cleanest precondition for moving the probability from “low” to “medium” is not a better narrative, but a purchase price returning to the report’s ideal range of ¥6,000–¥7,500. With the same earnings path and a starting valuation cut in half, the earnings/valuation conditions needed for “fivefold” immediately become much looser. This is the same sentence in growth-investing language: “good company ≠ good investment; the difference is price.”

  Jun 5, 2026
• QWhy has the market not realized all this yet? Is it too hard to understand, too easy to dismiss, or too far out? What could become the “narrative inflection point”?3/10

  A

  Bottom line: through Baillie Gifford’s “why has the market not realized this” lens, SCREEN is probably not a case of “too hard to understand / too easy to dismiss / too far out.” It has already been seen, understood, liked, and quite fully priced. The cognitive gap is weak, and may even lean toward overdrawn expectations. Fifteen analysts cover it, the average rating is Buy, and the 12-month target price is about ¥12,790, yet below the current price. It rose +17.94% in a single day on 6/3 to a new year-to-date high. This combination shows the market already trades it as a high-conviction long tied to AI equipment strength, advanced packaging volume, and FY2027 recovery. Baillie Gifford seeks the mismatch of “great but misread, still able to rise 5 times over the next 5–10 years.” SCREEN’s current issue is not that “the market has not realized how good it is”; it is that the market has realized it and prepaid the benefit. The report’s view that “the margin of safety is not obvious, close to none” is fully consistent with this. Positive expectation gap is not strong, so this question scores weak in the Baillie Gifford framework. Below, I separate “too hard to understand / too easy to dismiss / too far out,” then list the real possible narrative inflection points.

  ---

  1. It is not “too hard to understand.” The process barriers in cleaning equipment are deep, involving chemical-fluid management, particle control, yield qualification, and PoR status, but SCREEN’s selling points are already institutional consensus: global number-one share in single-wafer/batch/spin scrubber cleaning, TSMC contributing ¥89.7bn as a single customer, and advanced packaging increasing the importance of cleaning. These are all in annual reports and sell-side models. Coverage by 15 analysts is not the coverage level of an obscure stock. The market understands quite well what this company does and where its moat is. In the Baillie Gifford sense, “too hard to understand” requires complexity high enough for the mainstream to misread it, such as mistaking a platform business for hardware or network effects for cyclicality. SCREEN does not have this room for systematic misreading. It is a high-quality, strongly cyclical process-equipment leader, and it is priced as exactly that.

  2. It is not “too easy to dismiss.” A typical “dismissed” case is valued at low multiples as a sunset or ordinary cyclical stock, with the market refusing to assign a growth premium. SCREEN is the opposite: trailing P/E is about 27x, forward P/E ≈ 22x, and P/FCF ≈ 39.6x. The market is not applying a cyclical-stock discount; it is giving a growth-equipment-stock premium. The strongest bear argument in the report points directly to this: current high margins may partly be cyclical tailwind rather than pure structural monopoly, yet the market values it as a high-cycle growth stock rather than a medium-cycle equipment stock. In other words, the market does not dismiss SCREEN; if anything, it may overrate it slightly. The only angle that could barely count as “relatively dismissed” is that SCREEN is cheaper than Tokyo Electron at ≈ 49x, ASML at ≈ 56x, and KLA at ≈ 55x. But that is a reasonable discount for a narrower category moat and weaker indispensability. It is the deserved discount, not mispricing. “Relatively cheaper than top-tier leaders” does not mean “absolutely cheap,” and certainly does not amount to the kind of mismatch Baillie Gifford needs for a 5 times outcome.

  3. “Too far out” is barely discussable, but weak. The only dimension where a small mismatch may remain is that the market may still think in a cyclical framework and underestimate several structural long slopes: advanced packaging/chiplets/HBM moving cleaning and surface treatment from “auxiliary step” toward “process-critical,” after-sales recurring revenue from the installed base cushioning the cycle, and new products such as Lemotia PLP coating, LeVina direct imaging, and panel-level packaging cleaning opening non-traditional increments. If these turn into through-cycle share and margin steps, investors discounting it as a pure cyclical stock would be “not looking far enough.” But this line is weak for SCREEN for three reasons: ① much of its “far-out” story has already been prepaid in the 6/3 surge and 22x forward P/E. The market did not miss advanced packaging; it cheered and paid for advanced packaging. ② The true ceiling constraints, China exposure of 42.4% plus local substitution and partial wet-to-dry erosion, are downside far-out risks, not upside far-out opportunities, and this is negative under the Baillie Gifford lens. ③ Even under the most optimistic scenario, the report estimates only 7%–10% annualized over 10 years, far from Baillie Gifford’s “5 times,” or about 17%/year over 10 years. So the “too far out” layer has a narrow crack, but not enough to form the strong positive expectation gap Baillie Gifford needs.

  ---

  Why I rate this question “weak,” honestly: A true cognitive gap requires a verifiable upside mismatch between market pricing and long-term enterprise facts. SCREEN’s current state is high coverage, Buy ratings, a target price below the current price, implying slight downside, and a near-18% one-day surge on the AI/advanced packaging narrative. This is a stock where the narrative has already been told, the price has already reacted, and may even have front-run fundamentals slightly. It is not a stock whose narrative has yet to be discovered. Baillie Gifford’s firepower should concentrate on upside imagination in years 3–10, but SCREEN’s upside imagination has already been pulled forward into today’s share price. What remains is more a symmetric bet on delivery versus falsification than an asymmetric “the market will eventually wake up” opportunity. Without a strong positive expectation gap, this question is weak.

  ---

  Real possible “narrative inflection points,” which are re-rating triggers, not undiscovered value:

  • AI capex persistence is confirmed or falsified, the biggest variable. The 6/3 surge was essentially a market vote on a long upcycle in AI-driven fab capex. If TSMC/memory makers keep raising capex guidance for 2027 and beyond and cleaning-equipment orders deliver quarter by quarter, the narrative can upgrade from “cyclical strength” to “structural growth,” validating the valuation premium. That is an upside inflection point. Conversely, if AI capex enters a digestion phase or guidance is cut, the market will quickly reclassify SCREEN as a medium-cycle equipment stock. The report’s warning of a return to ¥5,000–¥7,000 and a 45%–60% drawdown is this downside inflection point. The current price is more heavily betting on the former, so downside odds are less favorable.

  • Advanced packaging cleaning truly scales, an upside inflection point that must be proven by revenue. If Lemotia PLP, LeVina, panel-level packaging, and other new products form a disclosed revenue curve in FY2027–FY2028 that is independent of the front-end wafer cycle, proving that the cleaning moat is widening into advanced packaging, the market may assign a higher-quality valuation that does not fall in sync with the front-end cycle. This is the most concrete positive catalyst today, but it still needs financial numbers to deliver. Before the numbers appear, it is only narrative.

  • Chinese local substitution is “confirmed” or “falsified,” a two-way factor and currently a sword hanging overhead. China accounts for about 40% of global fab-equipment procurement, and SCREEN China revenue is 42.4%. If local cleaning equipment makers clearly take orders at mature/mid-range nodes, that is a downside inflection point, with the core market being eroded. If high-end cleaning process barriers are proven difficult for domestic vendors to replace in the short term and SCREEN holds share, that removes one major discount item and becomes an upside inflection point. This is currently more a risk than an option.

  • Cyclical upturn confirmation, a delivery trigger rather than an expectation-gap trigger. SEMI expects 2026/2027 equipment sales of $139.0/$156.0 billion and 300mm equipment spending of +18%/+14%. If FY2027 revenue/profit turns positive year on year again and confirms that FY2026’s decline was only “halftime,” it will repair market concerns about a cycle peak. But this mostly validates expectations already priced in, rather than creating a new cognitive gap, so it has limited Baillie Gifford significance.

  One-sentence close: The market has not failed to understand, dismiss, or look far enough with SCREEN; it has seen it clearly, liked it, and paid in advance. The asymmetric upside mismatch Baillie Gifford wants is very weak here. At current levels, this is a symmetric bet whose direction depends on narrative inflection points and where downside odds are less favorable, not a growth stock where the market will eventually wake up. This confirms the report’s view of “good company, but not at a good enough price, with almost no margin of safety.” Putting it on a high-quality watch list and waiting for a better price is more honest than betting that “the market has not yet realized it.”

  Jun 5, 2026