Full Report

Figures converted from JPY at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

The Arena — Reclaimed and Prime Silicon Wafers

RS Technologies sits in the upstream materials layer of the semiconductor industry. Two products matter: prime silicon wafers (the polished discs that get processed into actual chips) and reclaimed wafers (used test/monitor wafers that are stripped, polished and resold for re-use). Prime is a five-player oligopoly worth roughly $13–16B globally; reclaim is a niche ~$700M sub-market that grows in lock-step with fab activity and is dominated by a handful of specialists. Understand those two markets, the chip cycle that drives them, and the position RST occupies inside both — and the rest of this report falls into place.

1. Industry in One Page

Takeaway: This is a supplier-of-suppliers business. Money is made by sitting upstream of every chip fab in the world and selling a consumable (prime wafers) or a recycled consumable (reclaimed wafers). Demand is set by wafer-fab activity, not by any single end product. The newcomer's mistake is treating it like a chip stock.

The industry has two distinct profit pools that share the same customer base:

• Prime wafers — virgin, polished single-crystal silicon discs (mostly 200mm and 300mm). About 90% of the world's IC output sits on the top-5 suppliers' wafers. The product is a commodity at the substrate level but qualification-locked at the fab level — a fab cannot swap suppliers without re-qualifying processes, which takes 6–18 months.
• Reclaimed wafers — wafers already used for testing, monitoring or process-development inside a fab, then re-polished and re-shipped back. Per RST's own annual report, about 20% of the wafers a fab consumes are test/monitor wafers, and roughly 80% of those are reclaimed rather than new — purely a cost decision (a reclaimed wafer sells at one-quarter to one-half the price of a new test wafer).

Customers are the same in both pools: foundries (TSMC, UMC, GlobalFoundries, SMIC), memory IDMs (Samsung, SK Hynix, Micron, Kioxia) and logic IDMs (Intel, Sony, Renesas, etc.). Profits exist because of high capital intensity, long qualification cycles, scale economics in crystal pulling and polishing, and — for reclaim — proprietary cleaning chemistry that extends the number of reclaim cycles per wafer.

Good cycles look like 2017–18 (smartphone + DRAM upcycle) and 2021–22 (post-COVID rebuild + AI server pre-build); bad cycles look like 2019 (memory glut) and 2023 (mature-node oversupply, China overbuild). Wafer demand lags fab capex by about 6–12 months and leads chip pricing by about the same.

2. How This Industry Makes Money

Takeaway: Both prime and reclaim are pay-per-wafer businesses with high fixed-cost plants, so cash flow is driven by capacity loading, not by chip pricing. Prime is more capital-intensive and lower-margin; reclaim is asset-lighter and structurally higher-margin because the input (a used wafer) is essentially free.

Pricing unit. Wafers are sold per piece (200mm or 300mm), often under multi-year supply agreements with volume bands and indexed pricing. Prime contracts are typically 2–4 years and include "take-or-pay" or minimum-volume clauses; reclaim is sold per wafer cycle with shorter commitments. Memory customers buy at the highest volumes and lowest unit prices; foundry/logic buy lower volumes but pay more for spec-tight 300mm.

Cost structure. Prime is dominated by silicon polycrystal feedstock (a Wacker / OCI / Tokuyama commodity), energy for crystal pulling, and depreciation on Czochralski pullers and polishing lines. Reclaim is dominated by chemistry (acids, slurries, filters), labor and depreciation — feedstock is the customer's own scrap. RST's reclaim segment runs 38–40% operating margins; prime runs 17–23% at the segment level. The 1,500–2,000 bps gap is structural, not cyclical.

Capital intensity. Building a new 300mm prime wafer line costs roughly $1B and takes 2–3 years to qualify with major foundries. A reclaim line is closer to $30–60M and qualifies in 6–12 months. That asymmetry explains why prime is a five-player game and reclaim has room for a handful of regional specialists.

Bargaining power. Customers concentrated (top 10 buyers ≈ 80% of demand); suppliers concentrated (top 5 prime makers ≈ 85% of 300mm capacity). The result is multi-year contracts with negotiated price floors — neither side has the upper hand in a balanced cycle, but in a glut customers can pressure pricing, and in a tight market suppliers can push through 5–15% annual increases (as happened in 2021–22).

The margin stack matters: reclaim is the cash cow, prime is the growth bet, and consumables is the synergy adjacency. Any read of this industry that lumps them together misses where the dollars actually come from.

3. Demand, Supply, and the Cycle

Takeaway: Wafer demand = (fab capex installed) × (utilization rate). When utilization drops below ~80%, wafer suppliers feel it within one quarter; reclaim feels it faster because customers cannibalize their internal monitor-wafer pools first.

Demand drivers. Wafer shipments scale with global IC unit demand, but the proximate driver is wafer-fab utilization — the share of installed capacity that fabs are actually running. Long-run secular drivers (AI servers, automotive electrification, edge inference, IoT) reach wafer makers indirectly, through fab capex one-to-two years later. Global silicon wafer surface area shipped has compounded ~4–5% over twenty years per SEMI, with sharp dips in 2009, 2019 and 2023.

Supply constraints. Prime wafer capacity additions take 2–3 years and are typically pre-sold under LTAs. The 2022–23 wave of Japanese, Korean and Taiwanese expansions added ~15–20% of global 300mm capacity, much of which is still ramping. Reclaim capacity is faster to add but bottlenecked by chemical-handling permits and skilled labor for polish-line tuning.

Cycle behavior. The downturn order is consistent: (1) memory price falls → (2) memory fabs cut utilization → (3) wafer LTAs flex down to volume floors → (4) prime ASPs slip 5–15% → (5) reclaim volumes hold up better (cost-savings tool for customers) but pricing softens. The recovery order is the reverse. RST's segment data shows this clearly: in FY2019 (downturn) reclaim margin actually expanded as customers re-used more wafers; in FY2022 (upcycle) prime wafer margin jumped from 17% to 26%.

4. Competitive Structure

Takeaway: Prime wafers is a five-player global oligopoly with no realistic new entrant in 200mm/300mm; reclaim is a fragmented but consolidating sub-industry where the top three players control roughly half the market. RST is the leader in reclaim and a sub-scale newcomer in prime — exactly opposite of every other player on this list.

Prime wafer concentration. Per Mordor Intelligence, the top 5 — Shin-Etsu Chemical (Japan), SUMCO (Japan), GlobalWafers (Taiwan), Siltronic (Germany), SK Siltron (Korea) — control ~85% of global 300mm capacity. Shin-Etsu and SUMCO together hold roughly half the 300mm market. GlobalWafers' 2022 attempt to acquire Siltronic was blocked by the German regulator, which is the clearest signal of how the EU views further consolidation at this level. China is the wildcard: GRINM/GRITEK (RST's JV partner) and several state-backed entrants are building domestic 200mm and 300mm capacity, but qualification with global foundries is still 3–5 years out for most.

Reclaim concentration. Per Market Research Reports, the global reclaim market was ~$662M in 2023, with the top three companies generating ~53% of revenue. RST's own annual report claims a 33% global share (cross-checked against the segment's $176M reclaim revenue against a ~$700M market = ~25% in USD terms; RST measures share in wafer units, where per-wafer pricing differences explain the higher unit share). The named competitor set — Kinik (Taiwan), Phoenix Silicon (Taiwan), Hamada Rectech, Mimasu Semiconductor (Japan), GST, Scientech, Pure Wafer, TOPCO — is mostly Japanese and Taiwanese, with no major Western player at scale.

RST's only true comp is Ferrotec — every other listed player is either a pure prime-wafer giant (Shin-Etsu, SUMCO, GlobalWafers, Siltronic) or a private reclaim specialist (Kinik, Phoenix, Mimasu). That uniqueness cuts both ways: scarce comp set means valuation is noisy, but no listed peer competes across all three of RST's segments at once.

5. Regulation, Technology, and Rules of the Game

Takeaway: Three external forces matter — China industrial policy (drives the build-out RST's Chinese JV captures), US export controls (limits what equipment Chinese fabs can buy and which technology nodes RST's Chinese subsidiary can sell to), and wafer-size transitions (12-inch displacing 8-inch displaces older reclaim revenue). Nothing else is decisive on a 3–5 year view.

The most important regulatory fact for RST is that its Chinese subsidiary GRITEK is a state-blessed national champion — listed on Shanghai STAR Market, capitalized in part by Dezhou City government, eligible for Big Fund III support. That changes the prime-wafer risk picture from "competing against $80B Shin-Etsu" to "competing inside a protected Chinese demand pool that Shin-Etsu cannot freely sell into."

6. The Metrics Professionals Watch

Takeaway: Wafer-industry analysts track six numbers that explain almost everything. Memorize these; you will see them in every transcript and presentation.

If you only watch three, watch SEMI MSI shipments, fab utilization (TSMC's quarterly disclosure is the cleanest), and wafer ASP commentary in SUMCO and GlobalWafers calls. Those three explain ~80% of every wafer stock's move on a 6-month horizon.

7. Where RS Technologies Co., Ltd. Fits

Takeaway: RST is a niche leader in reclaim stapled to a subscale challenger in prime stapled to a bolted-on materials portfolio. The reclaim leadership is durable, high-margin and global. The prime business is a leveraged China-only bet that depends entirely on Chinese fab build-out and is structurally lower-margin than the global oligopoly. The two should be valued separately.

The clean way to read RST: a market-share-leading reclaim company that has used its cash flow to plant flags in two adjacent businesses. Reclaim looks like a quality compounder. Prime looks like a venture bet inside a regulated emerging market. Consumables looks like a roll-up. Different segments, different multiples, different risks.

8. What to Watch First

Seven signals that quickly tell you whether the industry backdrop is improving or deteriorating for RST. Each is observable from public sources within 30 days of the underlying event.

(1) SEMI quarterly silicon wafer MSI shipments. If global wafer area shipped grows year-over-year for two consecutive quarters, RST's reclaim volumes follow with about one quarter's lag. Free, quarterly, and the single highest-signal data point for this name.

(2) Shin-Etsu and SUMCO quarterly commentary on 300mm pricing and LTA renewals. Both report on prime wafer pricing trajectory and LTA coverage. A renewed willingness by customers to sign multi-year LTAs at flat-to-up pricing is a strong leading indicator that the prime cycle has bottomed; cancellations or downward step-resets signal the opposite.

(3) TSMC and SMIC quarterly utilization disclosures. TSMC's overall and leading-edge utilization tells you about prime wafer demand at the top of the food chain. SMIC's utilization tells you about the Chinese mature-node demand that GRITEK actually serves.

(4) US Bureau of Industry & Security (BIS) export-control updates. Any new restriction on equipment or wafer sales to Chinese fabs at specific nodes directly resets the addressable market for RST's Chinese prime wafer subsidiary. The October-2022, October-2023 and December-2024 updates each materially changed the demand pool for sub-leading-edge wafers in China.

(5) Inventory days at the top-5 prime wafer makers. A 10–20% rise in days of inventory is the earliest reliable read on a downturn. It typically leads ASP cuts by 1–2 quarters, which then lead RST's reclaim and prime segment margins by another 1–2 quarters.

(6) RST's own segment shipment volumes and capacity-utilization commentary in the quarterly briefing decks. The company gives explicit colour on each segment's shipment trend and pricing — read the briefing PDF, not just the headline numbers.

(7) Chinese semi industrial-policy announcements (Big Fund disbursements, GRITEK customer wins). Each Chinese state-fund disbursement that backs domestic wafer customers tightens or widens the demand pool RST's prime business sits inside. GRITEK's own STAR Market disclosures (quarterly) carry the cleanest signal.

Know the Business — RS Technologies (3445)

Figures converted from JPY at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

RS Technologies is a three-businesses-in-one trench coat: a globally dominant reclaim-wafer specialist (37% margin, ~36% of revenue, ~71% of segment operating income), a sub-scale Chinese prime-wafer challenger that is China-only and subsidy-supported, and a bolted-on equipment / camera-module / battery rollup running at 5% margin. The market keeps treating consolidated FY2025 numbers as one cyclical wafer story; the reclaim engine inside is far higher quality than the blended P&L suggests, and the prime-wafer arm is far more China-policy-dependent than the segment label implies.

The right question is not "what's RST worth at a wafer-cycle multiple," but "what would each of these three businesses be worth standing on its own, and where is consolidated reporting hiding value or risk."

1. How This Business Actually Works

Takeaway: Reclaim is a near-defensive chemistry-and-polish business whose feedstock is essentially free (a customer's used wafer). Prime is a capital-heavy Chinese fab supplier whose feedstock is polysilicon and whose margin is tied to Chinese fab utilization. Equipment is a thin-margin rollup. The economic engine is reclaim, not wafers in aggregate.

A foundry like TSMC or a memory IDM like Samsung needs three kinds of wafers: prime wafers that become chips, plus test wafers and monitor wafers (the ~20% of wafer consumption that polices the process). Around 80% of those test wafers are recycled rather than bought new, because a reclaimed wafer sells for one-quarter to one-half of a new test wafer. RST takes the customer's scrap, strips off prior films with proprietary chemistry, re-polishes it to spec, and sells it back. Feedstock is free. The whole margin is in chemistry, polish allowance, and how many times you can recycle a single wafer.

The other two businesses run on different mechanics. Prime wafers (GRITEK in China for 8-inch power-semi wafers, plus a 39%-owned equity-method affiliate SGRS that is still pre-commercial in 12-inch) buy polysilicon, pull crystal ingots, slice and polish virgin substrates, and sell into mostly-Chinese fabs at competitive prices with state support. The Equipment & Materials segment is a roll-up: DG Technologies (silicon consumables for dry-etch tools), Union Electronics (trading/agency), RSPDH (optical pickup and in-car camera modules — acquired 2024, generated a one-off negative goodwill gain), and LE System (vanadium-redox-flow battery electrolyte). FY2025's revenue jump in this segment was almost entirely RSPDH's first full year of consolidation.

Reclaim barely flexes through cycles — 35.1% in the worst year (2020 COVID disruption), 40.6% at the 2022 peak. Prime swings 11.9% → 26.4% over the same window. Equipment trades at single digits.

Where incremental profit comes from. When Sanbongi and Tainan are at full utilization (which they were all four quarters of FY2025, per the Q4 briefing), every additional wafer slot adds revenue at near-segment margin because feedstock is free and the variable cost is mostly chemistry and labor. That is why management is bringing Sanbongi Plant 7 forward a year (now FY2026, was FY2027) and adding 70K wafers/month of new monthly reclaim capacity in 2026 — they have customer LTAs pre-signed and they cannot ship enough today.

Where margin gets pressured. Prime-wafer ASP is now the swing factor. In FY2025 8-inch shipment volume rose 20% YoY but average ASP fell about 10% YoY (general-purpose IGBT and memory wafers in China are the weak slice; power-semi wafers held up). Equipment margin is structurally low because optical-pickup modules and trading-company revenue have nowhere near reclaim economics.

2. The Playing Field

Takeaway: RST has no clean listed peer. The four prime-wafer giants are bigger, more cyclical, and operate the part of RST that is least profitable. Ferrotec is the only listed company whose mix even loosely resembles RST. The peer table shows RST cheap on multiples — but every peer is dominated by prime wafers, while RST's earnings power lives in reclaim. The "cheapness" is partly mismeasurement.

Peer figures sourced from Yahoo Finance quote-summary 2026-05-15 (TTM revenue and EBITDA), reported here in USD via Frankfurter/ECB rates. RST market cap from ¥6,640 close × 26.46M shares = ¥175.7bn ≈ $1,108M; EV is market cap less net cash of ~$485M; EBITDA is operating income $91M + D&A $35M = $126M. Multiples (EV/Sales, EV/EBITDA, EBITDA margin) are unitless and unchanged across currencies.

Three things the peer set reveals:

(1) RST is the cheapest in the group on EV/Sales (1.3x) and EV/EBITDA (5.0x), but cheapness here is not yet a mispricing. Half its revenue is the low-quality Equipment segment, which deserves an EV/Sales closer to 1x, and the Prime segment is at a Chinese-policy multiple. The pure reclaim engine is what could re-rate the stock; it is currently buried inside a blended multiple.

(2) Reclaim has no listed pure-play. Mimasu Semiconductor and the Taiwanese reclaim names are either too small or too diversified for a clean peer. Ferrotec is the only relevant comp — and it trades at 2.1x EV/Sales / 10.9x EV/EBITDA with worse mix (reclaim is only a slice of its business) and 13% segment-level op margins. Carving RST's reclaim out standalone would allow marking it against semi-materials specialists, not against SUMCO and Siltronic.

(3) The prime-wafer peer set shows what good and bad look like. Shin-Etsu and SUMCO together set 300mm wafer pricing globally and earn 9–28% operating margins depending on cycle position. RST's prime business runs 20% — credible inside the Chinese market but with no realistic path to global 300mm scale. GlobalWafers' 28x EV/EBITDA tells you what equity markets will pay for a clean prime-wafer pure-play in an upcycle — and how much that multiple can reset down when shipments slow.

3. Is This Business Cyclical?

Takeaway: Reclaim is the closest thing in semis to a defensive segment — in the 2019 memory glut and 2020 COVID year, reclaim margin actually rose (38–38%). Prime is fully cyclical. The blended P&L looks moderately cyclical; the right way to read RST through a downturn is reclaim flat-to-up, prime down 1,000 bps, equipment unchanged at thin margins.

The mechanism is counterintuitive. In a downturn, fabs cut wafer starts → they cannibalize their internal pool of test wafers more aggressively → they actually increase the share of test wafers that are reclaimed (because reclaim is cheaper than buying new test wafers). RST's reclaim volume holds; pricing is sticky because chemistry is qualified per customer. Meanwhile prime-wafer customers slow purchases, ASP softens 5–15%, and prime-segment op margin compresses by ~1,000 bps.

Two patterns the chart makes visible:

(a) Reclaim margin range FY2017–FY2025: 35.1% → 40.6%, a 5.5-point band. That is a quality semiconductor-materials margin, narrower in range than any IC manufacturer. The high point was the 2022 AI/memory peak; the trough was 2020 COVID. In the 2023 mature-node downturn — when SUMCO and Siltronic saw EBITDA fall double digits — RST's reclaim segment dipped only 100 bps from FY2022.

(b) Prime margin range FY2018–FY2025: 11.9% → 26.4%, a 14.5-point band. That is a fully cyclical wafer business. The 11.9% trough in FY2020 was driven by Chinese 8-inch IGBT ASP weakness and capacity ramp inefficiency; the 26.4% peak in FY2022 was the AI-server pre-build. FY2025's 19.9% with 8-inch ASP down ~10% YoY suggests we are mid-cycle on the down-leg.

The other cycle exposure is working capital and capex. RST's gross cash swung from $515M in FY2022 → $543M in FY2024 → $617M at FY2025 year-end while long-term debt rose to $90M because management is pre-funding the ~$359M capex plan for 2026–2028; net cash (cash less total debt) was ~$485M at FY2025 year-end. The business throws off enough cash ($95M operating CF FY2025, $47M FCF) to self-fund, but the next three years' growth depends on a customer demand forecast that has to materialize — if it doesn't, the new Sanbongi Plant 7 / Tainan Plant 2 / China lines come up at low utilization.

4. The Metrics That Actually Matter

Takeaway: Ignore consolidated revenue growth. The five things that drive value here are reclaim segment margin, reclaim capacity utilization, 8-inch prime ASP, the SGRS 12-inch equity-method loss, and ROIC trajectory. If you only watch one, watch reclaim segment operating margin — that's the one the consolidated number is most likely to obscure.

Relative health scored 0–1 within RST's own range — 1.0 marks the peak reading for that metric over the window.

Notice what the table is not showing: revenue growth and gross margin. Headline revenue grew 30% YoY in FY2025 — almost entirely RSPDH's first full-year consolidation (camera modules), which is the lowest-quality revenue in the mix. Gross margin compressed 200 bps for the same reason. Neither number tells you whether the business is getting better or worse. Reclaim margin and ROIC do.

5. What Is This Business Worth?

Takeaway: Consolidated multiples mismeasure this company. The right lens is sum-of-the-parts because the three segments have wildly different margins, growth profiles, capital intensity, geographic exposure, and acquisition multiples. The reclaim engine deserves a quality-materials multiple. The prime business deserves a China-policy multiple. The equipment business deserves a thin-margin diversified-industrial multiple. There is also an embedded listed stake — GRITEK on the Shanghai STAR Market (688521.SH) — that complicates the picture and should be sized before forming any consolidated valuation view.

The practical implication for an underwriter: when you see a "5x EV/EBITDA" headline for RST, ask immediately whose EBITDA. If you mentally split the FY2025 $91M segment EBIT into $65M (reclaim) + $27M (prime) + $10M (equipment) − $11M (other), and apply roughly 12x / 6x / 5x respectively, the math lands at ~$780M + $160M + $50M − $55M ≈ $925M of operating-business value, plus $485M of net cash, ≈ $1,410M of equity. At ~$42 per share the market cap is ~$1,121M. That suggests a ~25% gap to a clean SOTP — roughly half the reclaim multiple the market is not paying for, and roughly half the net cash that is earmarked but not yet deployed. It is also why this name's biggest swing factor is not next quarter's prime ASP but whether reclaim margins hold above 38% through the FY2026–28 build-out.

6. What I'd Tell a Young Analyst

The trap with RST is that consolidated metrics blend three businesses you would never value the same way. Read every press release with that filter. When net sales jumped 30% in FY2025, almost all of it was the RSPDH camera-module roll-on — the lowest-quality slice of revenue and the part most likely to fade once the negative-goodwill year is behind it. When ROIC fell from 12.7% to 10.8%, that's the capex pre-build phase, not a structural deterioration — wait for FY2028 readings before judging.

Watch four things, in order. (1) Reclaim segment operating margin quarterly — the company discloses it; if it stays above 38% while Sanbongi Plant 7 ramps, the quality-compounder thesis is intact. (2) 8-inch prime ASP commentary — management calls out the gap between power-semi pricing (firm) and IGBT/memory-grade pricing (weak), and this is the leading indicator for prime segment margin. (3) Sanbongi Plant 7 + Tainan Plant 2 capex disbursement — $109M Sanbongi + $156M Tainan committed over 2026–28; slippage or scope creep is the first sign customer demand has softened. (4) GRITEK's standalone Shanghai disclosures (688521.SH) — quarterly results there are the cleanest read on the prime segment without RST's HQ-allocated overhead distorting it.

The thesis breaks if: reclaim margin cracks below 35% structurally (chemistry cost-up, customer reclaim-cycle counts shifting), Sanbongi Plant 7 ramps at low utilization (demand miss), or US-China escalation forces GRITEK to write down its China prime book. The thesis works if: reclaim margin holds at 37–39% through the build-out, FY2027–28 revenue clears ~$660M with mid-teens operating margin, and management starts buying back stock when net cash crosses $500M again. Everything else is noise.

The mid-term plan published in February 2026 — $529M / $662M / $725M revenue and $97M / $110M / $120M op income for 2026 / 2027 / 2028 — implies operating margin grinding down from 18.6% to 16.5% as the rollup grows faster than reclaim. Track whether the mix actually plays out that way. If reclaim outgrows the plan, op margin should beat. If equipment outgrows the plan, op margin meets the plan but the multiple does not change. Same revenue, different P&L quality, totally different equity value.

Long-Term Thesis — RS Technologies (3445)

Figures converted from JPY at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

1. Long-Term Thesis in One Page

The 5-to-10-year thesis works only if the reclaim wafer engine compounds: shipment volume grows from 690K wafers/month today to ~1.5M+ by 2030 at a sustained 35–40% segment operating margin, the prime-wafer adventure in China matures into either a self-funding 12-inch business or a clean monetization, and management redirects the $485M cash pile into the moat rather than into another adjacency rollup. The durable value is not the consolidated P&L the market currently underwrites; it is one segment, two plants (Sanbongi + Tainan), one proprietary chemistry, and a customer base (TSMC, Samsung, SK Hynix, Intel) that does not switch reclaim suppliers in two-week sprints. The compounding math is plausible — a 9-year reclaim margin band of 35.1%–40.6% has already absorbed a memory glut and a COVID shock — but it is not yet evident in returns: ROIC has fallen from 19.2% (FY22) to 10.8% (FY25) and management itself authored a four-year walk-down to 16.5% consolidated operating margin by FY2028. The thesis decays into a portfolio bet if reclaim margin slips below 35% structurally, if the founder-CEO's ~43.89% combined voting block (8.04% direct + 35.85% via R.S. Tech Hong Kong Ltd) is used to favour the China complex over Japanese minorities, or if the next $359M of growth capex lands at sub-mature utilization.

2. The 5-to-10-Year Underwriting Map

What has to be true, what is true today, and what would break each leg.

Driver #1 is doing the most work. Every other line either flows from reclaim margin durability or is a hedge against the part of the business that does not earn its multiple. If reclaim margin holds the 35%+ floor through the FY2026–28 build-out, drivers 2 and 5 follow almost mechanically; if it does not, drivers 3, 4 and 6 cannot rescue the equity. The asymmetry is what makes this a one-question thesis.

3. Compounding Path

The compounding math is not heroic. It assumes (a) reclaim capacity grows ~70% by FY2028 in line with management's published schedule, (b) reclaim segment margin stays 36–38% through the ramp, (c) prime stabilizes at low-teens-to-high-teens margins through cycle, and (d) equipment stays at thin margins as a portfolio drag. The compound is in reclaim revenue × margin × time, not in multiple expansion.

FY2026–28 values per company mid-term plan disclosed Feb 2026 (converted at period-end FX rates from data/company.json.fx_rates); FY2030 extrapolated at the same per-wafer revenue and 37% margin assumption. Capacity figure includes Sanbongi + Tainan + Inner Mongolia footprint; reclaim revenue inferred from segment trajectory at constant ASP and disclosed cycle counts.

The chart makes the central tension visible: management's own FY2028 plan has operating margin still below FY2024. The bull case is not that management beats its plan on the consolidated line — it is that reclaim mix grows faster than guided, the equipment rollup matures or is rationalized, and consolidated margin re-expands toward 22%+ by 2030 as ramp dilution dissipates. Whether that happens is a function of capex discipline and capital allocation, not industry conditions.

4. Durability and Moat Tests

Each test is something the next 3–7 years will resolve in observable, recurring data.

Two tests are competitive (counter-cyclical margin survival; share gains vs SEMI MSI); two are financial (ROIC re-build; capital returns); one is customer-structural. The single most informative observation across the five is the next cyclical downturn — it is the unambiguous moat read, because the FY19 and FY20 reads were before the current capacity build-out and before the Ferrotec/Big-Fund-III pressure stack assembled.

5. Management and Capital Allocation Over a Cycle

The credibility split is unusually clean and unusually consequential. Promises made about the reclaim chassis — capacity additions, dividend trajectory, GRITEK STAR Market listing, plant timing — have been kept or beaten. Promises made about adjacencies — LE System scale, 12-inch SGRS commercial timing, the Feb 2024 "Upside Plan" — have been quietly retired without formal acknowledgement. A long-term shareholder must decide whether that pattern repeats with the next $359M of growth capex.

Founder Nagayoshi Ho controls ~43.89% of votes (8.04% direct + 35.85% via R.S. Tech Hong Kong Ltd) and runs every Chinese operating subsidiary personally. The alignment story is real — $485M personal stake against $1.4M total annual pay (365× ratio), no insider selling, options being exercised not cashed in. The structural concern is equally real: no designated successor, no separate chair, the China-relationship continuity at GRITEK and SGRS is concentrated in one individual, and the HK vehicle is a structural channel any future preferential capital transaction would flow through. The audit committee — 100% outside, CPA chair, China-qualified lawyer — is the minority shareholders' principal structural safeguard. For a 5-to-10-year position, the CEO succession announcement is a discrete event that either tightens or loosens the underwriting confidence; right now it is unanswered.

The pattern most worth tracking is whether the next mid-term plan (Feb 2027) reaffirms or withdraws the FY28 $724M / $120M numbers. Three walked-back plans in three years would collapse adjacency credibility entirely and reduce the thesis to a pure reclaim-segment underwriting — which still works, but at a materially lower equity value than current consolidated multiples imply.

6. Failure Modes

The thesis-breakers. Each is observable in public disclosures within 12 months of the underlying event.

7. What To Watch Over Years, Not Just Quarters

Five observable milestones across a 3-to-7-year window. Each updates the long-term thesis more than any single quarterly print.

The list is deliberately not a catalyst calendar. None of these resolve in a single quarter; they resolve over multi-quarter trajectories observable in segment disclosures, GRITEK STAR Market filings, and the Feb mid-term-plan refresh cycle. The Q1 FY2026 reclaim margin print (August 2026) is a near-term evidence marker that informs the trajectory — not the trajectory itself. A long-term shareholder should weight the 5-year reclaim margin shape and the next downturn read far more than any individual quarter.

The long-term thesis updates most if reclaim segment operating margin holds 36–38% through both the Sanbongi Plant 7 commissioning ramp and the next cyclical downturn — that combination would confirm the moat is structural, support a re-segmentation of the model toward quality-materials multiples, and engage the SOTP gap that has stayed unclosed for nine years.

Competition — Who Can Hurt RS Technologies, and Who It Can Beat

Figures converted from JPY at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

Competitive Bottom Line

RS Technologies has a real moat in one segment and a borrowed one in the other two. Reclaim is a global #1 (~33% share) at 36.9% segment operating margin, above 35% through every cycle for nine years — that combination of share, margin durability and a chemistry-and-qualification barrier is a genuine economic moat. Prime wafer and the equipment rollup are not moated; they ride on (a) Chinese state support for GRITEK and (b) acquired-niche margins that are average at best.

The competitor that matters most is not Shin-Etsu, SUMCO, GlobalWafers or Siltronic — RST is not in their league in prime wafers. It is Ferrotec (6890), the only listed name whose mix straddles reclaim, silicon parts and quartz crucibles the way RST's portfolio does, and the one most able to take reclaim share if pricing or qualification cycles open up.

The Right Peer Set

The peer set has to be split because RST competes in two distinct profit pools with completely different structures.

For the prime wafer business, the peers are the global top-5 oligopoly — Shin-Etsu Chemical (4063), SUMCO (3436), GlobalWafers (6488.TWO), Siltronic (WAF.DE), SK Siltron (Korea, unlisted). RST competes with none of them in 200mm/300mm prime wafers at global scale; its prime business (GRITEK in China + the 39%-owned SGRS 12-inch JV) sells almost entirely into the Chinese mature-node demand pool that the top-5 cannot freely serve under US export controls. They are the right benchmark for what good prime-wafer economics look like — and for what RST's prime segment is not.

For the reclaim wafer business, the only listed comparable with material reclaim exposure is Ferrotec Holdings (6890). Mimasu Semiconductor (8155), Kinik (1560.TW), Phoenix Silicon (6164.TWO) and the various private reclaim shops (Pure Wafer, Optim Wafer Services, KST World) are either too small, too fragmented, or not publicly disclosed enough for a clean comp.

Market cap and EV translated to USD at FX rates as of 2026-05-15 (frankfurter.app/ECB; TWD at static $0.0330/TWD). RST: ¥6,640 × 26.46M shares = $1,108M mkt cap; net cash $479M; EV ≈ $628M includes minorities and other adjustments. RST EBITDA is segment EBIT $90M + D&A $35M ≈ $125M. Source: data/competition/peer_valuations.json (Yahoo quoteSummary 2026-05-15).

Three reads of this peer chart:

RST sits in the bottom-left — cheap on EV/Sales (1.3x), middle-of-the-pack on EBITDA margin (25.8%). The cheapness is consistent with the consolidated mix (only ~36% of revenue is the high-margin reclaim segment; the rest is China-prime and the camera-module rollup). The 25.8% EBITDA margin understates the reclaim segment's true earning power because it blends three businesses with very different economics — see Where The Company Wins.

Ferrotec is the most relevant comp — same country, same diversified-materials mix, also has reclaim and silicon parts exposure, also has Chinese growth optionality. Trading at 2.1x EV/Sales and 10.9x EV/EBITDA on a 19.6% blended EBITDA margin. RST is cheaper on EV/Sales but slightly more expensive on segment-margin-adjusted multiples once you treat the camera-module rollup correctly.

The prime-wafer giants are not peers of RST in any meaningful sense — Shin-Etsu and SUMCO together set the global 300mm wafer price; GlobalWafers and Siltronic round out the global top-4. None of them compete with RST in reclaim, and none of them compete in China at the scale RST's GRITEK subsidiary does. They are the read-through for prime-wafer pricing and the ceiling for what a clean prime-wafer pure-play earns through-cycle — not direct competitors.

Where The Company Wins

Four advantages, each grounded in disclosed evidence.

The reclaim margin durability point is worth seeing visually — it is the strongest single piece of evidence that this is not just a cyclical wafer name.

SUMCO and Siltronic figures are full-group operating margins as a proxy (both are essentially pure-play prime wafer); reclaim segment data from RST FY2024 Briefing 12-year segment table + FY2025 results. The point is the shape: RST reclaim stays in a 5.5-point band while pure-play prime margins traverse 25+ point ranges and dip into losses.

Where Competitors Are Better

RST is sub-scale in the prime wafer market and has structural disadvantages there. Five concrete weaknesses, by competitor.

The most important of these is #3 — customer diversification in prime. RST's prime wafer revenue is structurally hostage to Chinese mature-node demand because GRITEK cannot freely export to the US and EU, and because SGRS's 12-inch product is pre-commercial. The China-only exposure earns the prime segment a discounted multiple regardless of segment margin — which is why our valuation framework values that segment 5-8x EBIT versus 10-14x for the reclaim segment.

Threat Map

The threats to RST's competitive position split into competitor moves, substitute technologies, and regulatory shifts. Each is dated and sized for severity.

The highest-severity threat is Chinese state-backed new entrants competing with GRITEK for the same captive demand pool. This simultaneously caps the prime segment's terminal multiple and creates the largest realistic downside scenario — Chinese 8-inch / entry-level 12-inch oversupply would compress prime ASP further than the −10% YoY already booked in FY2025. The reclaim moat is far more durable than the prime moat.

Threat scoring (0–1) — Severity if it happens, Probability of happening, Imminence of impact.

Moat Watchpoints

Five measurable signals an investor should track quarterly to know whether the competitive position is improving or weakening. Each is observable in public filings or industry data within 30 days of the underlying event.

(1) Reclaim segment operating margin (RST Q briefing). The single best read on whether the moat is intact. Above 38% is healthy and consistent with the 2017-2024 average; below 35% structurally would indicate chemistry costs are rising, customer reclaim cycle counts are improving (i.e. they're using their own internal pool more), or RST is conceding price at LTA renewals. The number is disclosed every quarter in the results briefing PDF. Currently 36.9% (FY2025) — the lowest in three years; watch FY2026 trajectory carefully.

(2) RST reclaim wafer shipment volume vs SEMI silicon MSI shipments. Calculate RST's reclaim revenue growth as a ratio to the industry's wafer-area-shipped growth (SEMI quarterly release). A widening gap in RST's favor means market share is gaining; a narrowing or inverted gap means share is leaking — most likely to Ferrotec, Mimasu, or Kinik. RST claims 33% share; this is the closest public proxy for verifying it.

(3) Customer concentration / new customer wins in reclaim. Watch RST's top-customer disclosure annually (in the yuho risk-factor section). If revenue is moving toward more diversified customer base, the moat is broadening. If concentrating into a single fab cluster (e.g. Taiwan only), the moat is narrowing. Equally important: track named customer wins or losses cited in any RST press release or briefing.

(4) 8-inch prime wafer ASP trajectory at GRITEK + SGRS 12-inch qualification milestones. GRITEK is separately listed on Shanghai STAR Market (688521.SH) and discloses quarterly results. Its standalone numbers are the cleanest read on the prime segment without consolidation noise. ASP YoY change and any commentary on customer qualification for the 12-inch product are the key data points. FY2025: 8-inch volume +20% YoY but ASP −10% YoY — that mix is unsustainable into FY2026 if Chinese new entrants ramp.

(5) Ferrotec semi-equipment segment growth and capex direction. Ferrotec's FY3/24 results showed the semi-equipment + Other segment growing volume but compressing margin (12.5% vs 18.2% prior year), with heavy capex into Malaysia and Changshan. If Ferrotec's reclaim-adjacent capacity additions materialize at 200mm or 300mm, RST's reclaim segment ASP and utilization are the first metrics to feel it. Track Ferrotec's quarterly briefing decks.

Current Setup & Catalysts — RS Technologies (3445)

Figures converted from JPY at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

1. Current Setup in One Page

The stock sits at an all-time high after a +150% one-year run, and the market is watching one thing above everything else: does reclaim segment operating margin hold 37%+ as Sanbongi Plant 7 begins mass production in FY2026? Q1 FY12/26 (reported 14 May 2026) printed reclaim margin at 38.5% (+2.5pt YoY) on full Sanbongi/Tainan utilization, the strongest live evidence the moat is broadening rather than eroding — yet management held FY2026 guidance unchanged ($529M revenue, $97M OP) and the only published sell-side target on the tape ($30 Minkabu) sits 30% below market. The near-term calendar is thin but high-signal: one hard event in the next 90 days (Q2 FY26 release, 7 Aug 2026), continuous Sanbongi/Tainan utilization reads, and an SGRS 12-inch slip already publicly disclosed for the next two years. The decisive multi-year update is the Feb 2027 mid-term plan refresh — a third consecutive walk-back of the FY28 $725M / $120M numbers would collapse adjacency credibility entirely, while a reaffirmation alongside the same reclaim-margin shape would support a re-segmentation of the model.

Recent setup rating: Bullish but extended.

2. What Changed in the Last 3–6 Months

The narrative arc has tightened. Six months ago the debate was "is RST a slow-compounding wafer cyclical or a hidden specialty-materials franchise?" After the February 2026 mid-term plan walk-down and the March-May rally, the live debate is now narrower and more decision-relevant: "is the +150% rally a justified moat-recognition trade, or a momentum overshoot ahead of the same FY28 margin compression management has authored in writing?" The Q1 FY26 reclaim print (38.5% on full capacity) is the first hard datapoint that tilts the live debate toward the bull mechanic — but the parabolic price action, the unrefreshed analyst marks, the SGRS 12-inch slip, and the founder's ~80% voting bloc are the four unresolved questions the next two prints will have to engage with.

3. What the Market Is Watching Now

The five live debates are not equal-weighted. Reclaim margin trajectory is the master watchpoint — every other line either flows from it or is a hedge against the parts of the business that do not earn their multiple. The market has already priced in continued reclaim strength via the +150% one-year move; the asymmetry from here tilts toward downside if reclaim margin compresses rather than upside if it merely holds.

4. Ranked Catalyst Timeline

The calendar concentrates the decision-relevant evidence between early August 2026 (Q2 FY26 print) and mid-February 2027 (FY26 full-year + new mid-term plan). Three of the ten items above are high-impact; the rest are useful context but unlikely to decide the underwriting on their own. The asymmetry sits in #1 and #3 — Q2 reclaim margin and the next plan refresh — because they update the two thesis variables (reclaim moat durability; adjacency credibility) that drive almost the entire equity value.

5. Impact Matrix

The matrix isolates the catalysts that actually resolve the underwriting debate from the ones that merely add information. Two near-term events (Q2 reclaim margin, Sanbongi 7 utilization) and two longer-horizon events (Feb 2027 plan refresh, SGRS/GRITEK trajectory) carry almost all the decision value. The buyback question is the swing factor for the alignment story but does not decide the moat. The RSPDH camera pivot is the cleanest "mostly noise" line on this page — it can disappoint without breaking the equity, and it can succeed without rescuing it.

6. Next 90 Days

7. What Would Change the View

The debate over the next six months turns on three observable signals, in this order of weight. First, the Q2 FY12/26 reclaim segment margin print on 7 August — a second consecutive quarter above 37% under live Sanbongi 7 ramp dilution supports a re-segmentation of the model toward quality-materials multiples and validates Long-Term Thesis driver #1; a print below 35% activates the bear's documented thesis-breaker. Second, the Sanbongi Plant 7 commissioning sequence through H2 FY26 — on-schedule mass-production start at ≥85% utilization confirms the customer LTAs that justified the pull-forward, while a slip or sub-70% commissioning fill rate refutes the moat-broadening narrative and re-escalates the capex/depreciation ratio. Third, the February 2027 mid-term plan refresh — a third consecutive walk-back of the FY28 $725M/$120M targets, or another SGRS 12-inch slip, would collapse adjacency credibility entirely, anchor the equity to a pure reclaim-segment SOTP at a materially lower multiple, and confine the bull case to "reclaim only." A clean reaffirmation paired with a first material buyback announcement would resolve both the moat durability question and the capital-allocation alignment question in one slide. Everything else on the calendar is texture; these three are the underwriting.

Figures converted from JPY at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

Bull and Bear

Verdict: Watchlist — the moat is real but the entry is poor, and one observable data point in three months sets the trajectory of the entire debate. Bull's sum-of-the-parts arithmetic is honest ($1.11B market cap vs ~$1.39B carved value) and the counter-cyclical evidence is the cleanest in listed wafers. Bear's case is equally honest: management itself authored a four-year margin walk-down to 16.5% by FY2028, the reclaim segment margin has slipped 370 bps over three years, and the prior mid-term plan was quietly retired. The whole debate collapses onto one shared fact — reclaim segment operating margin — and the first decisive read prints in August 2026. Paying $42 (RSI 87, +150% TTM, 49% above last published consensus) to sit through that read is poor risk control; the better course is to wait for the Q2 FY2026 segment table.

Bull Case

Bull's price target is $60 / share (~$1.58B market cap) over a 12–18 month window, built on a sum-of-the-parts on FY2027E earnings power: reclaim $72M EBIT × 13x, prime $28M × 6x, equipment $11M × 5x, less HQ overhead, plus $473M residual net cash. Cross-checks to ~11x FY27E EV/EBITDA, still inside the peer range. The disconfirming signal Bull names is reclaim segment operating margin below 33% for two consecutive quarters — the same metric the bear is tracking, just at a lower threshold.

Bear Case

Bear's downside target is $26 / share (-37% from $42) over 12–18 months, built from 6.0x EV/EBITDA on FY26E normalized EBITDA of ~$113M (stripping the $13M Gritek subsidy and the FY24-25 negative-goodwill flatter) plus $189M residual net cash after the FY26-28 capex draw, with a further 20% China-policy / governance / illiquidity discount on top. The cover signal is reclaim segment operating margin at 37%+ for two consecutive quarters while Sanbongi Plant 7 commissions on schedule above 85% utilization — i.e., evidence the FY25 slip was mix-and-ramp, not structural.

The Real Debate

Verdict

Watchlist. The two sides hold roughly equal weight on the durable question, with a tilt toward Bear at the current price. Bull's SOTP arithmetic is real and the counter-cyclical evidence is the cleanest in listed wafers, but the ~25% gap to fair value is too narrow to absorb a parabolic chart (+150% TTM, weekly RSI 87, 49% above the last published consensus mark) and management's own four-year margin walk-down. The single most important tension is the reclaim segment margin trajectory — the metric is already 370 bps off the FY22 peak, sits roughly 200 bps above the published moat-breaker, and the first decisive read prints in three months. Bull's read may hold because the FY25 slip is plausibly mix-and-ramp from RSPDH consolidation and Sanbongi commissioning; in that case a 37%+ Q1 FY26 print supports re-segmenting the model. The condition that shifts this to Lean Long is two consecutive reclaim segment margin prints at 37% or higher with Sanbongi Plant 7 commissioning above 85% utilization (durable thesis confirmation); a single Q1 FY26 print at 37%+ is a near-term evidence marker, not a thesis-breaker. The condition that shifts this to Lean Short / Avoid is two consecutive reclaim prints below 35%, which is the moat-breaker management itself effectively endorsed by guiding overall margin down to 16.5%.

Moat — RS Technologies (3445)

Figures converted from JPY at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

1. Moat in One Page

Conclusion: Narrow moat — segment-specific, not company-wide. RST owns one genuine economic moat (the reclaimed-wafer business, ~33% global share at 36.9% segment operating margin) and two segments that are essentially un-moated (Chinese prime wafers ride on state subsidy and an export-control ringfence; the equipment/materials roll-up earns 5% margin and competes on price). About 36% of FY2025 revenue and 71% of segment operating income live inside the moated segment. The rest of the consolidated entity is a portfolio bet, not a fortress.

A moat — used here in the Morningstar sense — is a durable, company-specific economic advantage that lets a business protect returns, margins, share or customer relationships better than peers. "Durable" means it survives downturns, technology shifts and management transitions. RST passes that test for reclaim and fails it for prime/equipment.

The strongest evidence is the shape of reclaim margins through cycles. Across FY2017–FY2025 the reclaim segment's operating margin moved in a tight 35.1%–40.6% band — a 5.5-point range. Over the same window, pure-play prime-wafer peers (Siltronic, SUMCO) traversed 25-plus-point operating-margin ranges and went into the red in FY2024–25. That gap is the moat showing up in numbers rather than in adjectives.

The weakest links are concentrated in the non-reclaim parts of the business: the prime-wafer segment lives entirely inside the Chinese mature-node demand pool (US export controls block sales into US/EU foundries), and FY2025 Chinese 8-inch ASPs fell roughly 10% YoY. The equipment/materials segment grew 87% YoY in revenue but earned only 5.3% margin — that is volume, not a moat.

Moat rating: Narrow moat. Weakest link: Reclaim ASP at LTA renewals + Ferrotec capacity adds.

2. Sources of Advantage

The candidate moat categories are tested below against actual disclosed evidence. Only two sources clear the "high proof" bar; three more are partial; and several commonly claimed advantages (network effects, brand, regulatory licence) do not apply to this business at all.

3. Evidence the Moat Works

A moat is only worth the data that confirms it. Six items support the reclaim-segment moat; two refute or qualify it. Both directions are shown.

The single most informative chart on this page: reclaim's margin shape vs the prime-wafer peer set over a full cycle.

Read the shape, not the absolute level. Reclaim stays in a 5.5-point band while prime-wafer pure-plays traverse 30 points and dip into operating losses. That is what durability looks like.

4. Where the Moat Is Weak or Unproven

Five places where the moat conclusion is fragile, soft, or borrowed. Investors should not treat the reclaim margin durability as a permanent fixture — the FY2025 reading is the lowest in three years, and the structural threats are stacking.

5. Moat vs Competitors

The peer comparison has to be split because RST's two profit pools have different competitor sets. Within reclaim there is essentially one relevant listed competitor (Ferrotec); within prime, four global giants set the benchmark for what good looks like — none of which RST competes with directly.

The peer table makes two points the consolidated multiple obscures:

(1) RST has the only reclaim-led moat in the listed universe. Ferrotec is the closest comp by mix but reclaim is a slice of its business; the four prime giants have wider moats but in a different segment. Mimasu and Kinik are sub-scale.

(2) Within prime, RST is sub-scale and borrowed. The 19.9% prime margin is comparable to peers' through-cycle averages, but the segment lacks LTA discipline, global customer reach and leading-edge qualifications. The right peer multiple for RST's prime segment is mid-single-digit EV/EBIT (China-policy-discounted), not the global peer set's 11–13x.

RST is the only data point with both high reclaim share and high best-segment margin; the prime giants sit at zero reclaim. Two different competitive ecosystems, one company straddling both.

6. Durability Under Stress

A moat that doesn't survive a real-world stress test isn't a moat — it's a tailwind. The reclaim moat has already been tested by the FY2019 memory glut and the FY2020 COVID shock, and it held. The harder tests are ahead: an aggressive Chinese new entrant, a US-China escalation, and the SiC/GaN substitution slope on a 5-year view.

Two stress cases have actual historical data on RST's response: the FY2019 memory glut and FY2020 COVID shock. The reclaim segment passed both — margin barely moved. The five untested cases — aggressive price war, China escalation, SiC substitution, multi-sourcing acceleration, cycle-count parity — are where the narrow-moat verdict could go either way. The single highest-information event for resolving the moat thesis in the next 12 months is FY2026 H1 reclaim segment operating margin: a 36–38% print confirms durability; a sub-33% print breaks it.

7. Where RS Technologies Fits

The moat is in one segment, two plants, one chemistry. Everything else is portfolio.

The bar chart makes the point in one frame: equipment is the largest revenue contributor but barely shows up on the income line; reclaim is mid-pack on revenue but dominates operating income. The investor who values RST on revenue multiples is paying for the wrong segment.

The fortress balance sheet matters less for moat valuation than is sometimes implied — net cash is optionality, not advantage. What it does is buy time: if FY2026–28 capex over-runs or reclaim margin slips, $485M of net cash means management is not forced into dilutive capital raises or distressed asset sales. That preserves the moat without enhancing it.

8. What to Watch

Eight signals that quickly tell an investor whether the moat is intact, broadening, or eroding. Each is observable in public disclosures within 90 days of the underlying event.

The first moat signal to watch is the FY2026 H1 reclaim-segment operating margin — if it holds at 36–38% as Sanbongi Plant 7 begins to ramp, the narrow-moat verdict is intact and the reclaim segment continues to earn a quality-materials multiple inside the consolidated entity. If it prints below 33% for two consecutive quarters, the moat thesis fails, the appropriate multiple resets to 6–7x EV/EBITDA on consolidated economics, and roughly a quarter of the equity value would be at risk.

Financial Shenanigans — RS Technologies (3445)

Figures converted from JPY at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

Forensic risk grade: Watch (38/100). RS Technologies's reported numbers tell a story that is broadly consistent with the underlying business, but three things keep this from a clean grade: (1) a gain on bargain purchase from the RSPDH acquisition flattered reported net income by roughly $9.6 million in FY2024 and $2.3 million in FY2025 — disclosed clearly, but it still inflates the headline EPS comparison; (2) the founder-CEO controls about 44% of shares through a Hong Kong holding vehicle plus a personal stake, with no independent counterweight on the executive side of the board; and (3) ordinary income leans increasingly on Chinese government subsidies routed through Gritek — $13.4 million in FY2025 — that management itself describes as partly non-recurring. Cash flow quality, accruals, and audit governance are clean. The grade would tighten to Clean if FY2026 net income hits the $63.8 million guide on a like-for-like basis without further negative goodwill or step-up subsidies. It would loosen to Elevated if receivables days creep above 130 again, the SGRS equity loss widens past $12 million, or the next acquisition produces another $6 million+ bargain-purchase gain through net income.

The 13-shenanigan scorecard

The scorecard concentrates the work. The income statement is broadly clean at the operating-income level. The ordinary-income line and net-income line are where the accounting judgment lives. Cash flow quality is genuinely good. Governance has structural concentration but disclosed processes.

Breeding Ground

The breeding ground is moderately elevated — founder concentration combined with M&A-driven growth is the dominant pattern. CEO Nagayoshi Ho (方 永義) controls 35.85% through R.S. TECH HONG KONG LIMITED and another 8.04% personally — roughly 44% combined. The audit-and-supervisory committee structure (監査等委員会設置会社) gives three outside directors a formal role, and KPMG AZSA (あずさ監査法人) audits the company. There is no public history of restatements, auditor resignations, or material weaknesses. The risk is structural rather than evidentiary: there is no independent challenge to the founder on the executive side of the board.

The breeding ground amplifies one specific concern: an M&A-heavy founder-led group that recently produced a $9.6 million gain on bargain purchase has the structural ingredients for accounting drift if discipline slips. The offset is that the audit framework and disclosure quality are functioning — the negative goodwill, the subsidy temporality, and the KPI definition change are all surfaced by management rather than uncovered by analysts.

Earnings Quality

Operating income is faithful; net income is flattered. Operating margin compressed from 22.1% (FY2024) to 18.6% (FY2025) — a real number, no aggressive capitalization gymnastics underneath. The distortion lives below the operating line: a gain on bargain purchase plus Chinese government subsidies make net income and ordinary income look smoother than the underlying trend.

The FY2024 receivables surge (+85% versus revenue +14%) is the single biggest reported red flag on the revenue-recognition test. The DSO chart below clarifies the cause: RSPDH was consolidated on December 31, 2024, so its full receivables balance hit the FY2024 closing balance sheet while only a sliver of its revenue showed up for the year. The FY2025 normalization to 106 DSO (still above the 89-day pre-RSPDH baseline, but well off the 144-day peak) supports management's explanation. If FY2026 DSO trends back to 90–100 days as RSPDH revenue annualizes, this concern fully resolves.

Gross margin has compressed 600bps from the FY2022 peak; operating margin is down 750bps. The compression is consistent with the disclosed mix shift: prime wafer 8-inch ASPs declined ~10% in FY2025, and the lower-margin semi-related/RSPDH segment grew 87% YoY. Headline net income looks resilient ($59.3M versus $60.2M) only because the FY2024 base contained a one-off bargain-purchase gain. This is not a manipulation finding — the income statement is honest — but the visible deceleration is masked by the comparison anchor.

Cash Flow Quality

Cash generation is the strongest signal in the file — clean, well above net income, not reliant on receivable sales or vendor stretch. Three-year CFO / net income runs at 1.58x. There is no evidence of factoring, securitization, or accounts-receivable sales. Vendor payments are not stretched abusively (DPO stable in the mid-60-day range). FCF is positive but modest — $27.0 million FY2024, $47.4 million FY2025 — because growth capex consumed roughly two-thirds of CFO during the recent investment cycle.

Working capital added roughly $23.5 million to FY2025 CFO — meaningful but not a lifeline. AR contracted in absolute terms (RSPDH stabilizing), inventory unwound $6.4 million, and accounts payable extended by $10.1 million. Ex–working capital, CFO would still be $71.2 million against $59.3 million of net income — the underlying conversion remains healthy. There is no sign of supplier-finance arrangements or one-time tax refunds boosting the cash-flow statement.

Metric Hygiene

Two metric-hygiene concerns and one definition change matter. Ordinary income is the metric Japanese investors anchor on, and it is materially flattered by $13.4 million of Chinese government subsidies plus $6.9 million of equity-method losses — the net non-operating contribution is roughly $15 million. Operating margin is the cleaner anchor.

The non-operating P&L tells a clean structural story. The FY2024 FX tailwind ($4.6M) reversed to a small FX loss in FY2025 ($0.6M). The Gritek subsidy doubled to $13.4M but is described by management as containing a past-period catch-up — the recurring run rate is probably closer to $6.4–7.7M. SGRS equity losses widened to $6.9M as the 12-inch China wafer build continues; this is a real cash burn through capital commitments, not just accounting.

What to Underwrite Next

This name belongs in the Watch bucket, not the position-sizing limiter or thesis-breaker buckets. The accounting is faithful at the operating-income level, and the non-operating items that distort net income are clearly disclosed. The forensic posture should be one of disciplined normalization, not skepticism — analyze the business on ex-negative-goodwill, ex-Gritek-subsidy ordinary income, and use operating cash flow rather than net income as the durability anchor.

Five things to monitor over the next four quarters:

1. FY2026 first-half DSO. Pre-RSPDH baseline was 89 days; FY2025 ended at 106. A trend back below 100 days confirms the FY2024 spike was acquisition timing. A drift above 130 days reopens the receivables question.
2. Gritek subsidy in FY2026 ordinary income. Management framed the FY2025 jump ($7.0M → $13.4M) as temporary. A reversion toward $6.4–7.7M would be consistent with that framing. A flat or higher number would mean the recurring subsidy stream is structurally larger than disclosed.
3. Negative goodwill in the next M&A. Management has spent $60M+ on M&A in the last three years (LE System, RSPDH, planned ~$232.9M over 2026-2028). If the next acquisition books another $6M+ gain on bargain purchase through net income, the pattern of opportunistically priced deals lifting headline EPS becomes a recurring framing concern.
4. SGRS equity loss trajectory. Equity-method losses widened from $4.4M to $6.9M as the 12-inch business in China builds. A widening past $12M in FY2026 would indicate the 39%-owned affiliate is consuming more capital than the equity stake suggests.
5. Operating margin recovery. FY2025 operating margin compressed 350bps. Management guides FY2026 to $535.9M revenue and $98.3M operating profit — implying 18.3% operating margin (vs 18.6% FY2025). The 5-year peak was 26.1% in FY2022. Whether margins stabilize above 18% or continue lower is the cleanest single read on whether the business is structurally maturing or in a cyclical trough.

What would change the grade. A clean FY2026 — $63.8M net income on a like-for-like basis, no further bargain-purchase gains, Gritek subsidy reverting toward $6.4M, DSO below 100 — tightens this to Clean (sub-20). A widening SGRS loss, a new acquisition with another $6M+ negative goodwill, or evidence of related-party flow that has not been transparently disclosed in the Japanese yuho moves the grade to Elevated (45-50). The grade does not become a position-sizing limiter unless one of two things happens: a regulator or auditor flag emerges (none today), or the founder-CEO uses related-party transactions to extract value from public shareholders.

Decisive paragraph. RS Technologies's accounting risk is a valuation-input adjustment, not a thesis breaker. The right normalized FY2025 net income is roughly $56.9 million (ex-negative-goodwill), not the reported $59.3 million — a 4% haircut to headline EPS. The right run-rate ordinary income is roughly $99 million (ex-Gritek subsidy step-up and ex-FX), not the reported $106 million — a 7% haircut. Applying those normalizations leaves screen multiples 4–8% higher than reported. That is the forensic insight that should travel into the valuation model. Position-sizing should respect the founder concentration (44% promoter stake leaves limited float discipline) and the China subsidiary complexity (currency, government, and minority-listing entanglements), but those are governance and structural risks, not accounting deception.

Figures converted from JPY at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

Governance: B — A Founder-Controlled Wafer Reclaimer With Clean Plumbing

A 55-year-old founder (born Oct 1970) holds 43.89% of the shares and runs both the Japanese parent and every Chinese subsidiary; that is the dominant fact. The board, the audit committee, and the related-party plumbing are unusually clean for a Japanese small-cap of this shape — but minority shareholders are along for whatever ride CEO Nagayoshi Ho decides to take, and the next governance test will be how the listed Shanghai subsidiary (GRITEK) is dealt with as he raises his stake there using parent capital.

The People Running This Company

Nagayoshi Ho (方永義) — Founded predecessor trading company Eikei Shoji (永輝商事) in 1999, set up RS Technologies in December 2010, and today sits as Chairman/Representative Director of every Chinese operating entity in the group: SDGRITEK (Shandong), Beijing GRINM RS, GRITEK (Shanghai STAR-listed), and the Taiwan venture Eyrus Semi. He took the company from a one-product Sanbongi reclaim plant in 2010 to $489M revenue / $59M net income in FY2025. Founder/operator track record is real; concentration of authority is total.

Satoru Endo — Manufacturing director who joined from Rasa Industries in 2011. He is the operating spine of the reclaim plants (Sanbongi, Tainan) and from February 2026 also runs DG Technologies as President. Wafer reclaim is a yield-and-process business, and Endo is the man who runs the floor.

Issei Osawa — Followed CEO Ho from Eikei Shoji in 2012, made director in February 2025, and now runs the Other-segment portfolio: LE System (vanadium-redox-flow batteries), Union Electronics, and the two Aisi Energy entities in Shandong and Panzhihua. He is the CEO's operating proxy in the diversification businesses.

Kiyohide Tomatsu — Joined from regional Kiraboshi Bank in 2023, became director in March 2024 as Strategy & Admin GM. He is the closest thing to a CFO function on the board and the only inside director with meaningful banking-side credibility for refinancings.

The succession question is unanswered. There is no designated successor for Ho, no chairman role separated from CEO, and no formal nominating-committee chair other than Ho himself acting through the Nomination & Compensation Committee. For a company where the founder's networks (especially in China) drive new business, this is the single biggest people-risk.

What They Get Paid

Pay is not the story here, ownership is. CEO Ho's $1.36M total package is roughly 0.02% of the $59M he and the team produced in net income — and 60% of it is fixed salary, not pay-for-performance. By Japanese mid-cap norms this is unremarkable; by global standards it is laughably restrained for a founder with this much shareholder value created. The 2023 introduction of a tightly-capped restricted-stock plan ($0.64M / 31,000-share annual cap, less than 0.12% of float) is a low-dilution, shareholder-friendly design. Outside directors receive ~$60k each — symbolic, not enough to dim independence, not enough to attract a marquee name either.

Are They Aligned?

Skin in the game: 8/10. CEO Ho's $490M (at ¥6,680/share, May 2026) stake is roughly 360× his annual cash pay. Other directors hold tokens, except Endo (137k shrs) and Osawa (125k shrs) who have meaningful but not life-changing positions. Foreign investors own 62.10% of the float, which gives the stock real liquidity but also means every governance signal — restricted stock grants, the Rasa cross-holding, China subsidiary capital injections — gets scrutinised by people who do not share the Japanese deference reflex.

Insider trading pattern. Japan does not publish Form-4-style individual trades, but the disclosed related-party history shows: CEO Ho's direct stake rose from 7.86% (FY2024) to 8.04% (FY2025) through stock-option exercises totaling $0.43M of strike consideration — i.e., the founder is converting options into shares, not selling. R.S.TECH Hong Kong Ltd has held its 35.85% block unchanged. No large-holder sales reported. The signal is "buying / accumulating," not selling.

Dilution is small and shrinking. The diluted-vs-basic EPS gap has compressed from 2.2% in FY2021 to under 0.4% in FY2025 as legacy 2019 stock-option tranches get exercised and the new RSU program comes in tight. The April 2026 RSU plan announcement (referenced in the yuho) is within the existing $0.64M / 31,000-share cap. Buybacks have not happened — the CFO said in the Feb 2026 briefing that buybacks would be "considered" subject to share-price, cash, and investment progress.

Related-party transactions are real but trivially small.

Total operating-level RPT exposure is $0.63M against $489M in group revenue — 0.13%, immaterial, and the disclosures explicitly state "市場実勢等を勘案して、一般的な取引条件で行っております" (terms based on market practice). The only judgment call is the $15.86M ($6.36M → $15.86M after a FY2025 +$0.45M top-up) strategic stake in Rasa Industries — Endo's former employer — held under the policy-investment rationale. This is not a serious red flag at this size, but it is the kind of position that a clean board should be ready to defend or sell.

Capital allocation. Investment is steered through the China complex: the 2025 capital increase at GRITEK (Shanghai-listed subsidiary) raised RS Tec's ownership in SGRS (Shandong) to ~39% via Gritek, and brought forward the restart of Sanbongi Plant No. 7. Dividend per share rose $0.22 → $0.29 (FY25) → $0.35 guided (FY26); payout is in the 13-14% range, leaving plenty of cash for the $255M reclaim capex plan. No buyback yet. Capital allocation is unambiguously growth-first.

Board Quality

Board Expertise / Independence Scorecard (1–10) — qualitative ranking based on FY2026 disclosure.

The Verdict

Governance Grade: B.

Grade: B. A founder-controlled Japanese small-cap with abnormally clean plumbing for its profile.

The strongest positives. Real founder ownership (43.89%) with multi-year option-exercise pattern that screams "accumulating," not selling. A 100%-outside audit committee staffed by a CPA and two lawyers — including a China-qualified lawyer at Nishimura & Asahi, which directly addresses the group's biggest operating risk. 100% attendance across all 16 board meetings and all 13 audit-committee meetings. Modest dilution overhang (under 0.4%) and a tight RSU cap. Related-party transactions exist but total under 0.15% of revenue. Pay is restrained and fixed-heavy — there is no LBO-style enrichment going on here. Big Four auditor (PwC) on a normal multi-year rotation with no off-balance non-audit consulting at the parent.

The real concerns. Concentration of authority: Ho is CEO and Chairman of every Chinese subsidiary that matters, including the Shanghai STAR-Market-listed GRITEK. There is no published successor plan. The board's strategy challenge is light — Nakano is a credible operator but adds aluminum-industry perspective, not deep semiconductor or capital-markets pushback. Cross-shareholding in Rasa Industries ($15.86M, increased in FY2025) is small but exactly the kind of position that needs to be defended explicitly each year. Female representation dropping back below 15% in FY2026 is going in the wrong direction. And the Hong Kong holding company structure for the founder's stake — while routine for cross-border Chinese-Japanese entrepreneurs — carries opacity that purely-domestic governance would not have.

Upgrade catalyst. A formal succession announcement (most likely Endo for operations + Tomatsu or a new hire for capital allocation) plus a buyback would push this to a B+. A separation of Chair and CEO roles, or a named lead independent director with override authority, would push it toward A-.

Downgrade catalyst. Any of three things: (1) a non-arm's-length transaction with the Hong Kong vehicle or with one of Ho's other Chinese ventures, (2) a take-private of the Shanghai-listed GRITEK subsidiary at terms favorable to the founder's complex over RS Tec minorities, or (3) a sharp expansion of the Rasa Industries stake without a clear strategic rationale. None of these have happened. All three are governance configurations the board structure as currently sized would struggle to challenge.

The one thing to watch. GRITEK is publicly listed in Shanghai and RS Tec just raised its stake via the January 2025 capital injection. How RS Tec's minority shareholders are treated as that subsidiary monetizes — versus how the Chinese strategic and Ho-controlled vehicles get treated — will be the next genuine test of whether this is a B that deserves to move higher, or a B that should have always been a C.

Figures converted from JPY at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

The Story — RS Technologies Co., Ltd. (3445)

The story since 2018 is one company in two halves: a quietly excellent reclaimed-wafer monopolist (Sanbongi/Tainan) bolted to a capital-hungry, cyclical, geopolitically exposed Chinese prime-wafer adventure (GRITEK/SGRS). Management's narrative has stayed remarkably consistent on what RST does, but the frame has cycled — from "comprehensive wafer manufacturer" (2018–2022), to "asset-value gap with a listed subsidiary" (2023), to a multi-engine conglomerate ("LE System + M&A upside", 2023–2024), and most recently to a "focused investment phase" (2025–2026) that quietly accepts margin compression and slips the China 12-inch ramp. The reclaimed-wafer chassis has delivered every promise made about it; almost every other promise — LE System scale, 12-inch prime ramp, the $930 million Upside Plan — has been walked back without ever being formally retired. Credibility on the core is high; credibility on the adjacencies is fading.

1. The Narrative Arc

Founder/CEO Nagayoshi Ho has been at the helm since the company was established on 10 December 2010. There has been no leadership transition. The current strategic chapter — RST as a "comprehensive wafer manufacturer" rather than a pure reclaimer — began in January 2018 when the Chinese prime-wafer maker GRITEK became a consolidated subsidiary. Every later twist (GRITEK STAR IPO 2022; LE System 2023; RSPDH 2024; holding-company transition disclosed Dec 2024) is a branch off that 2018 trunk.

The chart above frames everything else on this page. FY2018 is when revenue stepped up 2.3× — that is the GRITEK consolidation. FY2022 is when both revenue and margin peaked together — that is the GRITEK IPO year and the post-COVID semiconductor cycle peak. Every period after FY2022 has shown revenue continuing to grow while operating margin contracts. The new 2026–2028 plan formally extends this trajectory: management now guides FY2028 operating margin to 16.5%, down from 22.1% in FY2024 and 26.1% at the FY2022 peak.

2. What Management Emphasized — and Then Stopped Emphasizing

The reclaimed-wafer story stays bolted to every deck (33% global share, world No. 1). What rotates around it is more interesting: each year a "third engine of growth" is named, given a target, and either matured into reality (GRITEK) or quietly demoted (DG Technologies, LE System, the Upside Plan).

Missing required column(s): undefined not found in data set.

What rose: decoupling/regional strategy (1→4 from FY2021 to FY2023 as US-China friction crystallized), ROIC/ROE framework (introduced in FY2024 deck and now central), RSPDH (acquired Dec 2024 and immediately the largest new engine), AI demand for cutting-edge reclaimed wafers (first surfaced in Q4 FY2025 Q&A).

What faded: DG Technologies was the "third engine of growth" with three dedicated pages in the FY2021 and FY2022 decks; by FY2024 it is one bullet in the Semiconductor-related Equipment segment, and in FY2025 management admits it "did not contribute to operating income" and needs a new organizational structure. The "Upside Plan" — explicitly published in Feb 2024 with a $930 million FY2026 revenue target including $170 M from LE System and $305 M from M&A — was not mentioned in the FY2024 or FY2025 mid-term plan slides at all. It was replaced, not corrected.

3. Risk Evolution

The statutory yuho (Japanese securities report) carries roughly the same risk taxonomy every year, but the weight of individual risks moves visibly across the briefing decks, Q&A sessions, and CEO commentary. Three risks have grown materially since 2021; three have receded; one (Middle East logistics) is brand new in Q1 FY2026.

The two risks that have grown most loudly are 8-inch prime wafer ASP erosion (silent until FY2023, by FY2025 it accounts for the entire decline in prime wafer operating income: "the decrease in the unit price of some low-voltage wafers and general-purpose wafers such as IGBTs was noticeable") and capital efficiency below cost of capital. Management introduced the ROIC/ROE framework in February 2025 with confident claims that returns exceed CAPM-based cost of capital; one year later they conceded ROIC fell from 12.7% to 10.8% (vs. 9.0% WACC), accepted that "ROIC and ROE are on a downward trend," and asked investors to wait three years. 12-inch prime wafer ramp risk has also escalated — the equity-method investment loss from SGRS grew from $4.4 M (FY2024) to $6.4 M (FY2025), and the production schedule was officially slipped.

What has receded: the semiconductor cycle risk that defined the FY2023 commentary is now framed as past tense, with management arguing the reclaimed wafer business is "less susceptible to the silicon cycle" — a claim the 2023 prime-wafer numbers contradict for the prime side, but that the 40% segment margins in reclaimed largely vindicate.

4. How They Handled Bad News

The two biggest disappointments of the period were (a) the FY2023 prime-wafer cycle (segment sales –17.7%, operating income –37.6%) and (b) the FY2025 8-inch ASP collapse + 12-inch slippage. The handling of each is different and worth contrasting.

FY2023 — clean acknowledgement, framing remained credible. Management acknowledged the miss versus the FY2022 mid-term plan ("operating income decreased 8.6% year-on-year and –9.2% versus plan due to deterioration in the overall semiconductor market"), did not blame one-offs, and continued investing through the cycle. Subsequent reclaimed-wafer outperformance vindicated the framing.

FY2025 — quieter framing of structural margin compression. The FY2025 mid-term plan published Feb 2025 forecast FY2025 operating income of $96 M; actual came in at $91 M (–5%). The new mid-term plan published Feb 2026 cuts FY2026 OI to $97 M (vs. earlier $113 M path), with operating margin descending 22.1% → 18.6% → 18.3% → 16.7% → 16.5%. When asked directly on the Q4 FY2025 call whether "profitability will continue to decline until 3 years from now due to the large-scale capital investment," management answered: "Our group is currently in the investment phase. Therefore, we have included a plan to temporarily decrease the operating margin." The word temporarily is doing a lot of work in that sentence — the FY2028 endpoint of the new plan still has lower margins than FY2024.

5. Guidance Track Record

The table below filters out routine quarterly guidance and lists only promises that mattered for valuation, capital allocation, or strategic credibility. The pattern is unusually bimodal: every promise about the core wafer business and GRITEK has been delivered or beaten; almost every promise about adjacent businesses and capital efficiency has slipped.

Credibility = 6/10. Management deserves credit for the GRITEK IPO (a multi-year promise delivered on schedule and at a strong valuation), for honest handling of the FY2023 cyclical miss, for pulling Sanbongi Plant 7 forward, and for consistent dividend increases (¥10 → ¥55 per share over six years). The score is held back by a recurring habit: when adjacency targets miss, the next deck publishes lower numbers without acknowledging the prior promise. The Feb 2024 Upside Plan was the most visible example — an $836 M target gone in twelve months, never formally retired. The ROIC/ROE thresholds introduced with confidence in Feb 2025 were cut twelve months later. Taken together, the core deserves trust at face value; adjacency targets should be discounted ~30–40% by default.

6. What the Story Is Now

The story today is simpler than it has ever been, even though the consolidated entity is more complex. Reclaimed wafers are a high-quality, hard-to-displace cash engine running at ~37% segment margin and visibly growing. GRITEK 8-inch is mature but cyclical — useful, but already de-rated by the market. Everything else — SGRS 12-inch, LE System, RSPDH camera modules, the holdco restructuring — is option value that may or may not be worth what the consolidated capex line is costing.

De-risked over the period:

• Reclaimed wafer market position (33% global share held for six consecutive years, with Sanbongi/Tainan at full utilization and capacity stepping to 1.19 M/month by 2028)
• GRITEK listing and capital structure (executed cleanly; provided durable cushion of net cash plus listed-equity collateral)
• Decoupling architecture (US 4% of reclaimed shipments; prime sold mainly inside China; little direct tariff exposure)
• Balance sheet (net cash $468 M at end FY2025; DER 0.32x; investment-phase capex internally funded)

Still stretched:

• 12-inch prime wafer in China (publicly slipped multiple times; equity-method losses growing; "global environment and Japan-China relations" now openly cited as a reason)
• LE System / VRFB economics (FY2025 sales below $6 M vs. a former $170 M FY2026 plan; mass-production plant in Panzhihua scheduled for 2026 but commercial scale-up is unproven)
• Capital efficiency (ROIC has roughly halved from 19.2% in FY2022 to 10.8% in FY2025 and is now only ~180 bp above WACC; the plan accepts further dilution before recovery)
• RSPDH integration (optical pickup is a sunset business RST acquired for the manufacturing platform; the bull case requires automotive camera modules to scale in a market RST is openly describing as "highly competitive")

What to believe vs. discount:

• Believe at face value: reclaimed wafer capacity ramp, dividend trajectory, GRITEK 8-inch run-rate, and the company's reading of the tariff/decoupling environment.
• Discount by ~30–40%: the time-to-monetization on the 12-inch ramp, the FY2027–2028 revenue acceleration in the new plan (which presupposes RSPDH camera-module success and LE System monetization), and the implicit return-to-mid-20s operating margin after the "investment phase."

RS Technologies is not a turnaround or a transformation story; it is a high-quality core business whose owner has spent six years compounding net cash and capital intensity in roughly equal measure. The next three years will decide whether the diversification was capital efficiency or capital dilution. The evidence reads as inconclusive today — leaning toward dilution if the 12-inch and energy businesses don't show economics by FY2027.

Figures converted from Japanese yen at historical FX rates — see data/company.json.fx_rates for the rate table. Ratios, margins, and multiples are unitless and unchanged.

Financials — What the Numbers Say

1. Financials in One Page

RS Technologies is a structurally profitable, net-cash Japanese semiconductor-materials company whose reported financials hide two very different businesses: a high-margin reclaimed-wafer franchise that earns mid-to-high-30s operating margins, and a fast-growing but lower-margin prime-wafer subsidiary in China (consolidated since FY2024) that scales the top line and dilutes the group margin. FY2025 revenue jumped to $489M (+29.6%) while consolidated operating margin compressed to 18.6% (from 22.1%), free cash flow climbed to $47M, and net cash sits at roughly $485M against a market cap of about $1.11B — implying that more than 40% of the equity value is held in cash and securities, before any operating earnings are counted. The single financial metric that matters right now is the reclaim-wafer segment operating margin (FY2025: 36.9%): it is the cash engine, the moat proof, and the swing factor between an 8.7x EV/EBITDA semis-materials name and a low-teens-multiple cycle play.

2. Revenue, Margins, and Earnings Power

The starting point: a sub-$45M reclaimed-wafer business in 2013 has become a $489M semiconductor-materials group with three reporting segments. Growth came in two visible step-ups: the FY2018 spin-up of the prime-wafer subsidiary (GRINM, China) that more than doubled revenue, and the FY2024 onwards consolidation of additional Phenitec-related semiconductor equipment/materials business that pushed the third segment from $104M to $194M in a single year.

Revenue compounded at roughly 24% per year in USD over the 12-year window (lower than the yen CAGR because of yen depreciation), but the cadence is uneven: a strong 2017-2018 prime-wafer launch, a 2019-2020 cyclical air-pocket, a 2021-2022 boom, a 2023-2024 digestion phase, then the 2025 re-acceleration. Operating income tracks revenue with a one-to-two-year lag because heavy fixed-cost prime-wafer capacity comes online before it fills.

Gross margin has drifted from 38-40% (2017-2018, reclaim-dominant) to roughly 31% in FY2025, almost exactly the mix-shift you would expect as the lower-margin prime-wafer and equipment/materials segments grew faster than reclaim. The 2025 reset to 18.6% operating margin is not a quality break — the reclaim segment still earned 36.9% — it is a composition story. Whether the consolidated margin can re-expand from here depends on prime-wafer utilization in China and pricing in the semi-equipment segment, not on the reclaim business.

The quarterly view shows three things. First, the FY2025 step-up in revenue is real and broad — every quarter ran above $110M. Second, margins reset visibly in Q1-25 and Q1-26 (the seasonal pattern of lower Q1 margin is now built-in). Third, Q1 of FY2026 came in at $120M revenue, 19.0% operating margin — slightly ahead of the management guidance trajectory (full-year $527M / 18.3%) and supportive of the company's own three-year plan to $722M by FY2028.

3. Cash Flow and Earnings Quality

Net income tells you what the company earned on paper; free cash flow (FCF) tells you what it has left to compound — operating cash after the capital expenditure needed to keep the wafer fabs running. Until FY2023, RS Tech disclosed only annual capex and depreciation, so we can only fully reconcile the income/cash gap for the most recent two years.

Operating cash flow exceeded net income in both years, which is the right direction for an industrial business — depreciation and amortization ($27-35M/yr) bridges the gap. FCF lagged net income because of capex: $56M in FY2024, $47M in FY2025, against $27M and $35M of depreciation. The company is still in growth-capex mode — spending more than it depreciates — which is consistent with management's $722M FY2028 revenue plan but means FCF understates "steady-state" cash generation.

The 2014-2015 and 2019-2020 spikes correspond to the original Hokuto prime-wafer plant and the China prime-wafer expansions. Capex has run above depreciation in 9 of 13 years — the depreciation line is slowly catching up, which will mechanically raise the FCF conversion rate even if capex stays flat. A working approximation of "maintenance" capex is the current depreciation level ($35M); on that basis steady-state FCF would have been closer to $57-64M in FY2025.

4. Balance Sheet and Financial Resilience

The clearest piece of the RS Tech story is the balance sheet: $617M of cash against $132M of total debt at year-end FY2025 — net cash of roughly $485M, or more than 40% of the current market cap.

Three observations from the chart. First, the 2018 IPO/secondary materially recapitalized the company — equity went from $50M to $265M in one year. Second, the 2022 jump in cash and equity reflects strong operating earnings and the partial reorganization of the GRINM stake. Third, FY2025 saw debt re-emerge (from $60M to $132M) — long-term borrowings of $105M were drawn — but cash rose by an even larger $74M, so net cash actually improved.

The equity ratio is the eye-catching number: it collapsed from 82% at end-FY2023 to 38-39% in FY2024-25. This is not a leverage event — it is the consolidation of GRINM minority interest. About $467M of non-controlling interest now sits inside "net assets total" while the parent's own equity rose to $511M. On a parent-equity basis, debt/equity is still only 0.26x, and the current ratio of 4.3x is best-in-class. The balance sheet adds flexibility, not risk.

5. Returns, Reinvestment, and Capital Allocation

Return on equity dropped from ~38% in FY2017 to 6-8% during 2019-2023, then re-built to 12-14% in FY2024-25. The decline was not a profitability problem — operating margins were broadly stable — it was the IPO and subsequent earnings retention that grew the denominator. The pile of unused cash now embedded in the balance sheet (about $617M against $511M of parent equity) is the single biggest drag on the ROE number. A cash-adjusted return on operating capital — operating profit ÷ (assets ex. cash and goodwill) — would be in the 20-25% range and consistent with the moat narrative.

Dividends per share have risen from $0.08 (FY2015) to $0.29 (FY2025), with FY2026 guidance of $0.34 — roughly fourfold in USD terms over a decade (the yen path is steeper because of yen depreciation), but the payout ratio is still modest ($0.29 ÷ $2.23 EPS ≈ 13%). RS Tech has chosen retention over distribution: there have been no material buybacks, share count has crept up only slightly (26.37M → 26.46M over FY2024-25 from restricted-stock compensation), and most of the cash is being deployed into prime-wafer capacity in China and the LE System battery-electrolyte business. The implicit bet management is asking shareholders to accept is that reinvested cash earns more than the 13% dividend-equivalent payout would imply. The reclaim segment's 37% operating margin and the strong order book make that defensible, but the prime-wafer segment's mid-teen returns are doing less work for the share-count denominator.

6. Segment and Unit Economics

The single most important table on the page:

The economics live in reclaim. With about 36% of consolidated revenue, it generates roughly 71% of segment operating income ($65M of $91M before unallocated). Reclaim margins have run 35-40% for seven years, with only a modest cyclical dip in 2020. Prime wafer is meaningful in scale ($133M) but cyclical and capital-intensive — margins compressed from 26% in the 2022 boom to about 20% in FY2025. The semi-equipment segment grew 87% YoY but earns only 5.3% operating margin — almost a passthrough business at current scale. A reasonable read is that the stock is the reclaim business plus a cyclical option on prime wafer; semi-equipment is too low-margin today to materially move the equity story.

7. Valuation and Market Expectations

At $41.83 (close on 2026-05-15, converted from ¥6,640) with 26.46M shares, RS Tech has a market cap of roughly $1,107M. After backing out $485M of net cash, the enterprise value is about $622M on a pure cash-basis, or $1,102M if you keep the GRINM minority interest in (the more standard approach when the subsidiary is fully consolidated).

The market is paying ~19x trailing earnings, 8.7x EV/EBITDA, 2.25x EV/Revenue, 2.2x book value, and 23.6x trailing FCF. The 19x P/E sits roughly in line with kabutan's reported 17.7x (which uses a slightly different earnings denominator) and against a five-year average for the stock in the high-teens. Free cash yield is 4.2% on trailing FCF, 5-6% on normalized FCF if you assume capex moderates toward depreciation.

EPS doubled from FY2021 to FY2024 and has plateaued near $2.20-2.30 — the guided FY2026 EPS of $2.36 implies only 7-8% earnings growth, which the current 19x multiple already prices in. The market is paying for the mid-term plan: revenue to $722M and operating income to $119M by FY2028 (at current FX), which would put forward EV/EBITDA below 6x if reached.

Against today's $41.83, the base case implies modest upside (~9%) and the bull case roughly 40%. The bear case is the disciplined risk: if prime-wafer pricing weakens for a year and reclaim utilization slips, this becomes a low-teens P/E stock. Note the Stockopedia consensus target on record (¥4,443, about $28) was set when the share price was ¥3,805 (~$24) in early 2026 — the stock has since rallied to $41.83, blowing through that target. Analyst targets have not yet been refreshed publicly, so the market is now ahead of the published consensus.

8. Peer Financial Comparison

Three things stand out. First, RS Technologies trades at a discount to global prime-wafer peers on both EV/EBITDA (8.7x vs 11-28x) and EV/Sales (2.25x vs 2.5-7.8x). Second, that discount makes sense given (a) its much smaller scale, (b) the diluting effect of the lower-margin semi-equipment segment, and (c) the China-political risk premium attached to the GRINM subsidiary. Third, RS Tech's operating margin (18.6%) and ROE (12.5%) are competitive with the largest pure-plays despite the mix headwind — and the net-debt/EBITDA of essentially zero is the most conservative in the peer set after Shin-Etsu's net cash. On a fundamentals basis, the relative discount is wider than the quality gap justifies; the multiple should narrow as the FY2028 plan unfolds, if it unfolds.

9. What to Watch in the Financials

What the financials confirm. The reclaim-wafer franchise is high-quality (37% segment op margin, low capital intensity), the balance sheet is fortress-grade (net cash ≈ 43% of market cap), cash conversion is clean, and management's three-year plan is internally consistent with the current capex run-rate.

What they contradict. Consolidated returns on equity look mediocre (12-13%) because of unproductive cash on the balance sheet and the dilutive consolidation of a mid-margin Chinese subsidiary; the underlying economics are stronger than the reported numbers show. The 30% revenue jump in FY2025 was substantially helped by the consolidation of additional semi-equipment business — organic growth was closer to mid-teens.

The first financial metric to watch is the reclaim-wafer segment operating margin in FY2026 H1. If it holds at 36-38%, the moat is intact and the current multiple is defensible. If it drifts below 33% for two consecutive quarters, the story shifts toward a cyclical wafer name, the appropriate multiple resets toward 6-7x EV/EBITDA, and 25-30% of the equity value would be at risk.

Web Research — What the Internet Knows

Figures converted from JPY at historical FX rates — see data/company.json.fx_rates for the rate table. Ratios, margins, and multiples are unitless and unchanged.

The Bottom Line from the Web

External sources confirm an extraordinary repricing in process: the stock has compounded +85.6% in three months and +131.5% in one year to $42.08 (close 18-May-2026), with Q1 FY12/26 sales of $120M (+8.7% YoY) and operating profit $22.8M (+21.0%) coming in above IFIS consensus on every line. The single most thesis-relevant finding the filings underplay: management has silently slipped the China reclaim plant target from 150K wafers/month in 2026 to 50K in 2027 and 100K in 2028, citing "global environment and Japan-China relations" — a polite gloss on US BIS export-control creep that materially resets the China reclaim line; the SGRS 12-inch prime-wafer ramp is unchanged (110K→150K/m by 2026). Meanwhile founder/CEO Nagayoshi Ho controls ~43.89% of votes (8.04% direct + 35.85% via R.S. Tech Hong Kong Limited), which the market has tolerated because results keep beating but which still leaves minorities with limited recourse.

What Matters Most

1. SGRS 12-inch ramp pushed out — Japan-China relations cited as the reason.

2. Concentrated insider control: CEO owns ~43.89% of votes.

3. Stock has +131% in 12 months — a near-textbook semicap-Japan rotation winner.

4. FY2026 guidance kept, dividend hiked to $0.35 (+$0.06).

5. RSPDH (former Sony Precision Devices Huizhou) acquired for likely bargain-purchase gain; Sept-2025 Jiangxi Shinetech add-on pivots toward automotive cameras.

6. Chinese government subsidy to GRITEK of ~$13.4M in FY2025 — management calls part of it a "past-period catch-up."

7. Reclaim Wafer market: $662M (2023) → $1.09B by 2030 at 7.8% CAGR; RST is ~31% global share.

The Market Research Reports / SharedResearch.jp number-of-record on the addressable market is corroborated by company filings (top global share 31% per SharedResearch FY12/25 flash 14-May-2026; the company itself claims 33%). At RST's current FY25 reclaim revenues (~$191M segment scale per management commentary), the company already runs at multiple times its closest publicly tracked competitor (Mimasu Semiconductor 8155). (Source: marketresearchreports.com/blog/2024/09/02, sharedresearch.jp/en/companies/3445.)

8. Reclaim capacity build-out: 580K → 890K wafers/month by end-2026.

9. Inner Mongolia plant announced 19-Mar-2026 — adds a third regional reclaim/wafer footprint.

A 19-Mar-2026 disclosure (MT Newswires via MarketScreener) confirms RST will establish a silicon-wafer plant in Inner Mongolia. The news flow is consistent with the company's stated capex schedule (FY26-28) but adds geographic concentration risk inside China at exactly the moment US-China export controls are tightening on mature-node wafer flows. (Source: marketscreener.com/quote/stock/RS-TECHNOLOGIES-CO-LTD-20957314/.)

10. Outside director churn at March 2026 AGM — Shimizu out, Nakano in.

Recent News Timeline

What the Specialists Asked

Governance and People Signals

Compensation note. FY2024 disclosed pay for the CEO was $1.22M (66.2% salary, 33.8% bonus including stock and options). No other named executive has individual pay disclosed in the Yahoo Finance profile — typical of Japanese disclosure thresholds where only directors above ~$640K are individually itemized.

Insider transactions in last 6 months. No SEC Form 4 equivalents (Japan does not require US-style 5-day insider reporting at the same scale). The 24-Apr-2026 restricted-share issuance of $0.7M is the only material insider-grant event surfaced — small relative to the share base, plausibly retention-linked.

Industry Context

The reclaim wafer market grows mid-single digits over a long horizon; what makes RST exciting is not market growth but its already-dominant share (31-33%) inside a market that has just doubled-down on AI-driven test/monitor wafer demand. SEMI's Q1 2026 print of +13.1% YoY silicon wafer shipments to 3,275 MSI is the proximate sector tailwind. The downside-tail risks the search surfaced are concentrated in (1) US BIS export-control creep onto mature-node 200mm wafer flows (would directly compress Prime Wafer / GRITEK economics), (2) the silently delayed SGRS 12-inch program (which removes a major long-run revenue line until 2028), and (3) potential SiC/GaN substrate substitution shrinking long-run reclaim TAM. None of these are short-term thesis-breakers; all three matter for terminal value.

The market structure is consolidated. SUMCO (3436), Shin-Etsu Handotai, Siltronic, SK Siltron, GlobalWafers and Soitec dominate prime silicon; reclaim is a sliver of the value chain with RST as the unambiguous leader and small private/regional competitors (Mimasu, Kinik, Phoenix Silicon) holding the residual share. The competitive risk is not "share loss to a peer" but "TSMC/Kioxia/Samsung in-housing reclaim" — a scenario neither management nor specialists have addressed externally with hard customer-retention numbers.

Figures converted from JPY at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

Web Watch in One Page

At $41.83 the equity has absorbed roughly three-quarters of the bull sum-of-the-parts target on a +150% twelve-month rally, but the entire 5-to-10-year thesis collapses onto a small number of observable signals. These five watches track the variables that actually decide the underwriting — reclaim segment margin durability, capacity-execution credibility, the competitive moat under attack, the borrowed regulatory ringfence shielding GRITEK, and capital allocation discipline under a founder who controls 79.6% of the votes and has retired three mid-term plans in three years on the non-reclaim side.

The set is deliberately weighted toward evidence that would change the long-term view rather than evidence that anticipates the next quarterly print. The single near-term hard date — Q2 FY12/26 results on 7 August 2026 — sits inside watch #1; everything else either feeds, contextualizes, or counterweighs the reclaim-margin verdict. Cadences are tighter (1d) where the signal is event-driven and time-sensitive, and looser (1w) where the relevant disclosures are slow-moving governance, plan refreshes, or capital-return announcements.

Active Monitors

Why These Five

The report converges on five open questions that span five-to-ten years rather than five quarters, and the watch list mirrors them in order of weight:

1. Will the reclaim moat hold under Sanbongi 7 ramp dilution? This is the metric the entire equity sits on; if margin holds 36–38% the model re-segments toward quality-materials multiples, if it slips below 35% structurally the appropriate multiple resets to mid-single-digit EV/EBITDA.
2. Is the capacity build-out customer-driven? A clean Sanbongi 7 commissioning at high utilization converts the pull-forward into operational proof. A slip refutes the moat-broadening claim.
3. Is the moat actually being attacked, or is the threat narrative academic? Ferrotec capex pace and any Big Fund III disbursement decide whether the 35–40% margin band is structural or temporary.
4. Does the borrowed ringfence on prime survive the next BIS cycle? SGRS 12-inch has already slipped twice; a third slip combined with a new mature-node rule collapses $139–277M of remaining GRITEK capex into a stranded-asset risk.
5. Will the next mid-term plan and capital-return cycle confirm or refute the adjacency-credibility discount the variant view applies? The Feb 2027 plan, a first buyback, and any related-party transaction through the founder's HK vehicle are the discrete tests that resolve the adjacency-premium debate the +150% rally has not yet absorbed.

The watch list is intentionally light on calendar-driven catalysts (sell-side initiations, AGM cycle, sector data points like SEMI MSI) because those tend to inform without resolving. Each of the five above is structured so that a single disclosure can change the underwriting view — either confirming the bull SOTP toward $59.85, validating the bear's $26.46 downside, or holding the verdict in the current watchlist posture until the Q2 FY26 segment table prints.

Figures converted from JPY at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

Variant Perception — Where We Disagree With the Market

At $41.83 the market has absorbed roughly three-quarters of the bull sum-of-the-parts target while the only published sell-side mark sits at $29.62 — a +150% twelve-month rally that the analyst community has not refreshed. That gap is the consensus signal: the price is doing the underwriting, and it is implicitly trusting the FY2028 $725M / $120M mid-term plan at face value. Where we disagree is narrower and more decision-useful than the bull/bear arguments already in the deck. The rally embeds an unearned adjacency premium, the prime segment's borrowed regulatory ringfence is being priced as a delay when it is actually a structural ratchet, and the ordinary-income headline that Japanese retail anchors on is flattered by $6–10M of non-recurring items that the multiple does not reflect. Each is observable in segment disclosures within the next 3–9 months.

Variant Perception Scorecard

A 64 score reflects two things. First, the strongest of our three views (the adjacency-credibility discount) is materially different from the observed market price but only partly different from the report's bear case; the variant lives in pricing, not in raw analysis. Second, consensus clarity is moderate because only one named analyst publishes a live target (Minkabu/IFIS $29.62) and the market price has decoupled from it — the rally itself is the cleanest revealed consensus. Evidence strength is high: every claim below traces to a disclosed segment table, a transcript quote, or an externally documented policy event.

Consensus Map

What the market appears to believe, where we can see it, and the assumption embedded in each view.

The most important row is the first one. Coverage is structurally thin — Simply Wall St counts four analysts (Daiwa, Ichiyoshi, Okasan, plus one), only one of which publishes a live target — so the price has become the consensus signal. At $41.83 against a Bull SOTP of $59.85 and a Bear target of $26.46, the market has priced ~80% of the bull-case path. That is where the variant view earns its keep.

The Disagreement Ledger

Three ranked disagreements. Each names a specific market assumption, the evidence that breaks it, and the metric that resolves it.

Disagreement #1 — the adjacency premium is unearned. Consensus would say "management has guided FY28 $725M / $120M, the reclaim engine is broadening, treat the plan as base case." Our evidence is the guidance scorecard: the same management retired the Feb 2024 Upside Plan ($867M FY26 → silently replaced with $529M FY26), the LE System $159M FY26 target (FY25 actual under $6M), and the SGRS 300K/m by 2027 milestone — all without a formal acknowledgement that they had been withdrawn. The right base case applies a 30–40% credibility discount to the non-reclaim portion of FY28 operating income. If we are right, the market would have to concede that $32–44M of FY28 plan EBIT is uncertain and the appropriate consolidated multiple is 6.5–7.5x EV/EBITDA on the un-discounted plan rather than 9x. The cleanest disconfirming signal is a Feb 2027 plan refresh that reaffirms FY28 numbers AND shows LE System above $19M run-rate AND announces an external OEM design win for the RSPDH Jiangxi Shinetech camera pivot.

Disagreement #2 — SGRS is a ratchet, not a delay. Consensus reads "150K → 50K wafers/month is a timing slip; the option value is preserved." Our reading is that this is the first observable signal that mature-node BIS controls will eventually capture wafer flows — every prior tightening cycle (Oct 2022 / Oct 2023 / Dec 2024) shows the same one-directional creep, and management's "Japan-China relations" framing is the gloss on a policy wedge that has only ever closed. The Inner Mongolia plant disclosure (19 Mar 2026) doubles down on Chinese geographic concentration at the precise moment the policy backdrop deteriorates. If we are right, the prime segment terminal value collapses from the Bull frame's ~$170M toward ~$63–95M, and the remaining $139–277M of GRITEK capex compounds the destruction. The cleanest refuting signal is GRITEK STAR Market (688521.SH) showing 8-inch ASP YoY narrowing below −5% and subsidy income normalizing around $9.5M annualized.

Disagreement #3 — the ordinary-income anchor is partially synthetic. The Japanese investor base anchors on ordinary income (経常利益), framed by management as "record high for five consecutive years." That headline at FY25 $105.9M embeds $13.4M of Chinese subsidies (which the CEO labelled partly a past-period catch-up) and a residual $2.6M negative-goodwill flatter on net income. Normalize, and the recurring run-rate is ~$96–99M — a 4–8% haircut. On its own this disagreement does not move the equity by more than ~$3/share, but it tightens the asymmetry created by disagreement #1: the FY28 $120M op-income target sits on top of an FY25 base that is itself slightly flattered.

Evidence That Changes the Odds

The seven items below are not generic facts — each one moves the probability of the variant view in a direction the consensus market price has not yet absorbed.

How This Gets Resolved

Six observable signals. Each has a current state, a print that validates the variant view, a print that refutes it, and a timing window. Resolution is a multi-quarter trajectory, not a single earnings event.

The signals are weighted unevenly. Signal #1 (Feb 2027 plan refresh) carries the bulk of the resolution because it directly tests management's adjacency-delivery track record over a fresh cycle. Signal #2 (reclaim margin trajectory) is the closest thing to a binary near-term test but only updates the reclaim leg, not the consolidated underwriting. Signals #3 and #6 (SGRS + BIS) move slowly but in one direction; #4 (buyback) tests the capital-allocation half of the alignment story; #5 (coverage refresh) is the mechanism by which a variant view becomes consensus.

What Would Make Us Wrong

The most honest red team is to take each of the three disagreements and name the evidence that would force us to retract it before the market does.

On the adjacency premium (#1). We are wrong if the Feb 2027 plan refresh is the first clean reaffirmation in three years — FY28 $725M / $120M held at face value, LE System cleanly above a $19M annual run-rate (it has spent FY25 at under $6M), and RSPDH camera modules either showing an external OEM design win or pivoting cleanly toward a recurring automotive revenue base. The Feb 2026 Sumitomo Electric LE System selection is one data point in the right direction; one more credible external commercial validation would weaken the bearish overlay materially. We are also wrong if the second mid-term plan in the 5-to-10-year window comes in matching the Feb 2026 plan within ±5% on operating income — that would convert "three retired plans" into "three plans then a stable plan," and the discount we are applying becomes recency bias.

On the SGRS ratchet (#2). We are wrong if BIS does not issue a new mature-node wafer rule within the next 12 months and GRITEK's STAR Market filings show 8-inch ASP YoY narrowing below −5%, and a first non-Chinese fab qualification for SGRS materializes (e.g., a Japan-headquartered foundry / IDM technology-transfer disclosure). The structural-ratchet read fails fastest if the policy backdrop holds and the operational read recovers — both have to break the same way for it to be merely a delay rather than a permanent reset. We are also wrong if SGRS proves capable of serving the Chinese mature-node demand pool at scale without non-Chinese qualifications — i.e., the addressable market is large enough that the "borrowed ringfence" is itself a durable franchise even with policy uncertainty.

On the ordinary-income flatter (#3). We are wrong if FY26 Gritek subsidy income comes in at or above $12.6M — that would invalidate the CEO's own "past-period catch-up" framing and mean the recurring subsidy stream is structurally larger than disclosed, which actually raises normalized ordinary income rather than lowering it. We are also wrong if the next M&A footnote shows no bargain-purchase gain — that would break the pattern we are pricing in. This is the lowest-conviction of the three views and the easiest to retract; it survives only as a tightening overlay on disagreement #1, not on its own.

The first thing to watch is the Q2 FY12/26 segment table on 7 August 2026 — not because it resolves the variant view (it does not), but because a 37%+ reclaim margin print under live Sanbongi 7 ramp dilution confirms the asymmetric trust split (reclaim delivers, adjacencies do not) that underpins disagreement #1 and tightens the bear-case asymmetry on the rest of the equity.

Liquidity & Technical

Figures converted from Japanese yen at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

The data file marks RS Technologies (3445) as institutionally tradable, size-aware — a $12.0M average daily traded value supports a 5% position for funds up to roughly $266M over five sessions at a 20% participation cap, but execution discipline matters because the median intraday range is 4.2%. Technically the stock is in a powerful uptrend (price 70% above the 200-day SMA, golden cross active since August 2025) that has just reached overbought extremes — RSI peaked at 87 a week ago and has rolled to 71, MACD histogram has shrunk from +240 to +99 in two weeks, and price is sitting at the all-time high.

Portfolio implementation verdict

Price snapshot

The 10-year price chart

Read on the long sweep, the chart shows three regimes: a 2017–18 first run ($5 → $30), a four-year sideways consolidation between $20 and $30, and a vertical breakout starting April 2026 that has carried the stock to a fresh all-time high in six weeks. The current move is a clean upside resolution of the 4-year base, with the 200-day curling decisively upward beneath price. (Note: yen-translated chart compresses the recent rally because the yen has weakened versus the dollar over the 2016–2026 window; the native yen chart shows the underlying move more cleanly.)

Relative strength vs benchmark

Momentum panel — RSI + MACD

RSI printed 87 on 8 May 2026 — the deepest overbought reading in the available history — and has already rolled to 71. MACD histogram peaked at +240 in the same week and has collapsed to +99. Both indicators are still positive, but the rate of change has turned: this is what late-stage momentum looks like, not early-stage thrust. Near-term (one-to-three month) bias is for a digestion phase rather than another vertical leg.

Volume, volatility, and sponsorship

The last three weeks show weekly volume roughly 60–95% above the 50-day baseline — the kind of sponsorship change you want to see confirming a breakout, not a parabolic blow-off on thin tape. The 20-day ADV (318k shares) is 40% higher than the 60-day ADV (227k shares), which says the supply/demand profile has stepped up with the price move.

Top historical volume spikes

Two of the three biggest volume spikes in 10 years sit in the immediate post-IPO window of August–September 2016 (the company listed August 2016), which is mechanical rather than catalyst-driven. The November 2022 spike paired an 11× volume day with a +5.5% gain and is the only "real" sponsorship event in this top-three slice — the recent April 2026 breakout is approaching but has not yet eclipsed those historical extremes on a single-day multiple basis.

Current realized vol is 66.7% — between the 80th percentile (59.3%) and the 10-year max (89.9%). That is the "stressed" band. A position sized off historical-average vol will under-budget risk; size off the current 67% reading or wait for vol to mean-revert to the 35–50% band before adding meaningfully.

Institutional liquidity panel

ADV has stepped up materially in the recent rally — 20-day is 40% above 60-day, and annual turnover of 134% means the float changes hands more than once a year. ADV-to-market-cap of 1.09% is healthy for a mid-cap Japanese semicap name; for context, anything under 0.5% would be the "specialist only" zone.

Fund-capacity table (reverse the math)

The middle column is the practical answer. A fund with $266M in AUM can build a full 5% position in five trading days at a 20% ADV participation cap; halve that to $133M if you want to be conservative at 10% ADV. A $630M global fund is capacity-constrained at a 5% target weight — it would need 2–3 weeks of build, and the same again on exit.

Liquidation runway by issuer-level position size

A 1% mcap position ($11M) clears in one trading week at 20% ADV participation and in two at the more disciplined 10%. 2% mcap is the boundary: nine days out at the aggressive cap, more than three weeks at the conservative one. Anything beyond 2% is a multi-month build/exit and should be sized as a specialist-only exposure.

Daily-range proxy: the median daily range over the last 60 sessions is 4.2% — well above the 2% threshold the spec flags as elevated impact cost. Use limit orders, work the position in tranches, and budget a 25–50 bps slippage cushion vs typical mid-cap execution.

The largest issuer-level position that clears in five trading days is 1.0% of market cap at 20% ADV and 0.5% at 10% ADV — that is the binding constraint on this name.

Technical scorecard + stance

Stance — bullish bias, extended on a 3–6 month horizon. The trend, sponsorship, and golden-cross structure are consistent with higher prices over the medium term; the near-term read is that momentum has gone parabolic, realized vol is in the stressed band, and the stock is bumping the all-time high. Not a chase setup. Watchpoint on pullback toward $34.65 (the 20-day SMA and Bollinger middle band, also the breakout pivot of late April) — a break of that level would call into question the new uptrend and warrant a reassessment. Watchpoint above $45.30 (the 52-week and all-time high, also the Bollinger upper band region) — a weekly close above $45.40 with continued volume confirmation would re-arm the bullish setup with blue-sky targets in the $54–57 zone as a referenced range. Liquidity is not the constraint for funds under ~$250M building a 5% weight; above that, capacity discipline implies a 10% ADV participation cap and a 2–3 week build window — manageable, but a real input to the entry plan.