NAURA Technology

Phase A — Understand the business

Lens 1 · Company Overview

NAURA Technology Group (北方华创) is China's largest wafer-fab-equipment (WFE) maker — the domestic analogue to Applied Materials, and the only Chinese company inside the global top-10 chip-tool vendors. It rose to #6 globally by WFE revenue in 2024 and #5 in 2025.

What it makes. NAURA is the broadest domestic tool portfolio by process step: plasma/dry etch, PVD and CVD deposition, oxidation/diffusion (thermal/furnace), cleaning, and anneal on the semiconductor side. Yole/analyst commentary calls this breadth "an exception" among Chinese vendors and dubs NAURA "China's Applied Materials". Segment-share estimates inside China: ~30% of etch, ~25% of thin-film deposition, and low-single-digit % of CMP — CMP domestically is led by Hwatsing, not NAURA.

Corporate structure. The listed entity runs four business lines: (1) semiconductor process equipment — the growth engine; (2) vacuum & thermal equipment; (3) lithium-battery / photovoltaic equipment; and (4) precision electronic components (resistors, crystal devices, modules). The semiconductor-equipment segment is the dominant driver: ~¥13B in 1H25 vs ¥21B+ across FY24, i.e. roughly 65–75% of total revenue and rising.

Founding & ownership. Established September 2001 by Beijing Electronics Holdings (a SASAC state entity) as Beijing Sevenstar Electronics; IPO 16 March 2010 on ChiNext. Acquired Beijing North Microelectronics (NMC) — the silicon-etch/PVD business — in 2016 and renamed to NAURA on 24 Feb 2017. It is state-controlled: Beijing Electronic Control (a subsidiary of Beijing State-owned Assets) is the actual controller with ~43.03% of shares. The National IC Fund ("Big Fund") led a ~$1.3B private placement.

Customers. Not disclosed by name in NAURA's public English materials, but the demand base is the Chinese leading-edge and mature-node fabs under localization mandate — SMIC, CXMT (DRAM), Hua Hong, YMTC, and the broader China-fab buildout. Contract structure is capital-equipment sales (tool + process recipe + aftermarket service), not recurring/SaaS; the recurring layer NAURA is building is aftermarket/spares monetization to lift margin mix.

One-line business model: sell the picks-and-shovels of China's chip-independence drive to a captive, policy-compelled domestic buyer base.

Lens 2 · Supply Chain

Upstream (inputs into NAURA): precision-machined chambers and mechanical modules; RF generators, mass-flow controllers, robotics/EFEM, vacuum pumps, and high-purity gas-delivery subsystems — historically the highest US/foreign-content bottleneck; specialty materials and control electronics. Under Entity-List presumption-of-denial (see Lens 10), NAURA has been localizing critical subsystems and bringing components in-house.

Midstream (NAURA): integrates these into etch / deposition / thermal / clean tools; ships tool + process recipe.

Downstream (buyers): China fabs — SMIC, CXMT, Hua Hong, YMTC and second-tier logic/power/analog fabs; end-market is Chinese-made ICs for domestic electronics, autos, and AI-adjacent silicon.

Named chokepoints / single-source dependencies:

• US subsystem content — RF power, precision valves/MFCs, certain vacuum and metrology subsystems have been US-sourced; the Entity List makes these presumption-of-denial. NAURA's mitigation is stockpiling + domestic substitution, but this is the live single-point-of-failure.
• Lithography adjacency — NAURA does not make litho scanners; it addressed the coat/develop (track) gap by taking control of Kingsemi (see Lens 3).
• Ion implant, metrology/inspection, and EUV/DUV scanners remain gaps NAURA cannot fill — these still depend on foreign tools (AMAT/Axcelis for implant, KLA for inspection, ASML for litho) or other domestic vendors.

Names present → lens satisfied. The chain's defining feature: NAURA is simultaneously a beneficiary of the sanctions (downstream demand) and a target of them (upstream inputs).

Lens 3 · Competitive Advantages (moats)

1. Breadth-as-moat / platform bundling. NAURA is the only domestic vendor spanning etch + deposition + thermal + clean at scale. That lets it bundle a multi-step process solution to a fab — a switching-cost and integration advantage no single-tool Chinese rival (Piotech in deposition, Hwatsing in CMP, ACM in clean) can match. This is the "China's AMAT" thesis in one sentence.
2. Policy moat / national champion. State control (~43% via Beijing Electronic Control), Big-Fund capital, and a government mandate for 70% local tooling by 2027 mean NAURA's addressable demand is administratively created, not competed for. Domestic-equipment adoption already beat its 2025 target, hitting 35%, led by NAURA + AMEC.
3. Installed base + process know-how. Every tool shipped seeds recipe lock-in and aftermarket pull-through — the compounding advantage NAURA is now trying to monetize.
4. Consolidation power. NAURA is the acquirer in China's equipment consolidation wave — Kingsemi (litho track) and Chengdu Guotai Vacuum (vacuum coating) fold capability in and remove would-be independents.

Bargaining power: Over customers — high and rising, because sanctioned fabs have shrinking non-NAURA options. Over suppliers — weak on the US-content pieces (presumption-of-denial), improving as localization proceeds. The moat is real but geopolitically contingent: it exists because Washington drew a line, and it would erode fastest if that line moved (unlikely) or if domestic rivals commoditize each process step (the live threat — see Lens 13).

Lens 4 · Segments

NAURA does not publish a clean English product/geography segment table, and the research-layer segments.csv is empty — so this is /, not ``.

By business line (approx.):

Geography: overwhelmingly domestic China — "China-driven procurement accounted for the majority of order intake in 2024–2025". Export exposure is minimal (and constrained the other way — NAURA can't easily sell into US-allied fabs).

Why the mix matters: the semi-equipment segment carries the growth and (relatively) the margin; the electronic-components line is a structural gross-margin dilutant, part of why blended GM sits below Western WFE peers (Lens 5). The mix is shifting toward semi-equipment (good) but within semi-equipment toward more competed process steps (margin-negative — Lens 13).

---

Phase B — Measure performance

Lens 5 · Earnings Result

Five-year P&L (CNY millions):

The FY2025 story is a margin story, not a growth story. Revenue compounded ~4x in four years and still grew +30.9% to ¥39.35B, but:

• Net income fell 1.8% YoY (¥5,622M → ¥5,522M) — the first down year in the run.
• Operating margin compressed ~680bps (21.6% → 14.8%) and gross margin ~300bps (42.5% → 39.5%).
• The proximate cause is R&D up +47% YoY (¥3,699M → ¥5,435M) as NAURA races to close the litho/implant/metrology gaps, plus mix drag from lower-margin components and price competition in core process tools.

Latest quarter — Q1 2026: revenue ¥10.32B (+25.8% YoY), net profit ¥1.63B (+3.42%), R&D +36.6% to ¥1.40B, operating cash flow +143.3%, EPS ¥2.2556. Order backlog scheduled through Q1 2027 — strong forward visibility.

Balance sheet flags: total assets ~¥91B; shareholders' equity ~¥39.3B; current ratio 2.43 (healthy); D/E ~34.6%; but negative net cash ~ −¥1.05B and periods of negative operating cash flow during the 1H25 build (the "cash strain" Digitimes flagged around the Kingsemi takeover). The Q1-2026 OCF surge (+143%) suggests the working-capital strain is easing.

Market reaction / what's priced: the stock is up massively over the window and trades above the mean analyst target (Lens 7) — the tape has treated even the entity-listing as bullish (forced localization). What is not yet priced is a margin trough deeper than the R&D bulge.

Conflict to flag: one data-provider quoted FY2025 operating margin ~21.7% (TTM) and net margin 22.5% — these are inconsistent with the stockanalysis line items (¥5,522M net on ¥39,353M rev = 14.0%, arithmetically fixed) and with Smartkarma's "net income declines to 5.52B amid a 31% revenue surge." I treat 14.0% net / 14.8% operating as correct and the ~22% figures as stale/mislabeled TTM cuts.

Lens 6 · Earnings Calls (sentiment trend)

NAURA holds Chinese-language results briefings, not English earnings calls; transcripts/ is empty. Sentiment is reconstructed from management framing in disclosures and trade press.

Tone arc, FY2024 → Q1 2026:

• FY2024 / early-2025: triumphant — "third consecutive year of growth," breakthroughs in "new etching and deposition products… boosting revenue and market share," #6 global rank. Management leaned into share-gain and product-breakthrough language.
• Mid-2025 (1H25): confident on volume but cash-and-margin caveats appear — Digitimes: "profit rises 15% amid cash strain, Kingsemi takeover". The story shifts from pure growth to investment intensity.
• Late-2025 / Q1-2026: the recurring new phrase is "domestic substitution" and "localization ratio"; the phrase quietly de-emphasized is margin expansion. Management now frames R&D intensity as the cost of closing gaps, and points to backlog through Q1-2027 as the visibility anchor.

What they keep saying: localization, domestic substitution, platform breadth, R&D breakthroughs. What they stopped saying: margin leadership. That shift is the signal — management is telling you the next phase is share-and-scale, paid for with margin.

Lens 7 · Comps

Peer WFE table. Multiples are with source/date or `n/a`; NAURA's own CSV and all peer CSVs in the research layer are empty, so nothing here is. Mixed listing currencies — market caps normalized to USD `` at spot.

Dividend yield / 5-yr avg ROE: NAURA proposed a ¥7.62 / 10 shares cash dividend on FY2025 → ~0.9% yield; 5-yr avg ROE n/a as a clean series (FY2025 ROE ≈ ¥5,522M ÷ ~¥39.3B equity ≈ 14% ).

The comps verdict is the whole valuation tension. NAURA trades at ~107x trailing / ~80x forward — a premium to ASML (65x), AMAT (60x), KLA (36x) — despite (a) being a smaller, lower-margin, single-country vendor, (b) being under a US presumption-of-denial license ban, and (c) posting a declining net income. The domestic peers tell the same story: AMEC ~31x sales, NAURA ~15x sales, versus US-listed ACMR at ~1.1x sales / 10x P/E. The A-share market is paying a national-champion / scarcity premium an offshore investor cannot underwrite on fundamentals alone.

Lens 8 · Stock-Price Catalysts (5-year, >5% moves)

Intraday tick data isn't in the research layer; catalysts are ``.

• Dec 2024 — Entity List addition (one of 140 Chinese firms). Counter-intuitively net-bullish for the shares: the market read forced localization as a demand accelerant. This is the single most important catalyst and it did not break the stock.
• Quarterly earnings prints — the reliable ±5% movers. Q1-2025 pre-announcement (+51% rev guide) and the FY2024 beat (+44% net) drove rallies.
• Kingsemi stake (10 Mar 2025) and subsequent control move (Jun 2025) — strategic-gap-closing catalysts, well-received.
• Global-ranking headlines (#6 in 2024, #5 in 2025) — sentiment catalysts amplifying the national-champion narrative.
• Policy prints — the "70% local tooling by 2027" mandate and "domestic adoption hit 35%" act as sector-wide re-rating catalysts.

Pattern: for NAURA the market reacts to (1) the localization policy clock and (2) revenue/backlog acceleration far more than to margin. Entity-List news is a demand signal here, not a risk signal — the mirror image of how the same news hits a US-facing chip name.

---

Phase C — Judge people & books

Lens 9 · Management

• Chairman: Zhao Jinrong. CEO: Tao Haihong. Both operate inside a state-controlled governance structure (Beijing Electronic Control ~43%). This is a state-appointed professional-manager archetype, not a founder-owner — capital allocation answers to Beijing's industrial policy as much as to minority shareholders.
• Track record: the team executed a genuine industrial build — 4x revenue in four years (¥9.7B→¥39.4B) and a climb from also-ran to #5 global WFE, while integrating NMC, Akrion (2018, US, CFIUS-approved), Kingsemi, and Chengdu Guotai. Operationally credible.
• Capital allocation: aggressive reinvestment + consolidation — R&D up to ¥5.4B (13.8% of sales), plus M&A to buy capability (Kingsemi ¥1.69B) — funded partly by Big-Fund private placement (~$1.3B). Modest dividend (~0.9% yield). ROE/ROIC trend is decelerating as the equity base swells faster than earnings (FY2025 ROE ~14% vs richer prior years) — the classic strategic-investment-phase profile.
• Skin in the game: management personal ownership is minimal (state-controlled float); alignment is policy alignment, not equity alignment. Insider-transaction data n/a (no insider-transactions.csv).
• Red flags (governance): (1) state control means minority interests can be subordinated to national objectives; (2) capital raises dilute (private placements to state funds); (3) heavy reliance on government subsidies in reported profit is characteristic of the Chinese semicap cohort but the specific subsidy quantum is n/a here (do not fabricate; flag for the shelf).

Net: competent state-champion operators executing an industrial-policy mandate well. Judge them on localization share and tool breakthroughs, not on per-share value creation — the incentive structure points at the former.

Lens 10 · Forensic Red Flags

grounding: web-only — there is no 10-K / 10-Q to forensically read and no SEC enforcement record possible (no CIK). This lens is therefore weaker than for a US filer; findings are / and flagged as such.

Accounting / quality-of-earnings watch items:

• Margin compression despite record revenue — GM 42.5%→39.5%, operating margin 21.6%→14.8% (FY24→FY25). Not fraud, but the tape's central quality signal: growth isn't converting to profit.
• Government subsidies in reported profit. Chinese semicap names typically book material government grants through other income; this can flatter net income and is not a durable earnings stream. Specific figure n/a — a required item to pull from the Chinese annual report before relying on net-margin trends.
• Working-capital / cash conversion. Reported negative net cash (~−¥1.05B) and prior negative operating cash flow during the 1H25 build. For a capital-equipment vendor, contract liabilities (customer advances) and inventory build are the lines to watch — advances should rise with backlog (good), but receivables outrunning revenue to state-directed customers would be the warning sign. Specific AR/inventory/contract-liability lines n/a here.
• R&D capitalization vs expensing — a common Chinese-GAAP lever (capitalizing development spend to protect margin). NAURA appears to be expensing heavily (R&D up +47%, hitting net income), which is conservative and credible — but the capitalized-vs-expensed split is n/a.
• Acquisition accounting — Kingsemi / Chengdu Guotai bring goodwill and consolidation effects; watch for goodwill and minority-interest complexity.

Regulatory findings (read regulatory/regulatory-findings.md, generated 2026-07-06):

• SEC (EDGAR LR + AAER): 0 findings — not applicable. NAURA has no CIK and is not an SEC filer; no EDGAR enforcement search is possible.
• US export-control action (material, non-SEC): NAURA was added to the US Commerce Department Entity List in December 2024 — a presumption-of-denial license requirement on all US-origin technology/components, including R&D collaboration. A predecessor/affiliate, Beijing Naura Magnetoelectric Technology, was on the Unverified List (Oct 2022). This is the single most consequential regulatory fact about the company — but it functions as a demand catalyst plus an input constraint, not a fraud/enforcement red flag.
• Non-SEC enforcement search ("NAURA Technology" FTC OR DOJ OR FDA OR CFPB OR settlement OR fine): no material fraud/enforcement hits surfaced beyond the export-control listing.
• Item 3 / Legal Proceedings: no US filing exists to quote; Chinese annual-report legal disclosures n/a here.

Verdict: No fraud/enforcement red flag. The genuine forensic issues are (a) earnings quality (subsidy dependence + margin compression) and (b) a hard input-side sanction — both real, both ``-grade, and both requiring the Chinese annual report to underwrite properly. Verified via SEC EDGAR EFTS (N/A — no CIK), web search, and export-control record as of 2026-07-06.

---

Phase D — Project & stress-test

Lens 11 · Forward Projection

Built bottom-up from FY2025 actuals + FY2026 consensus. All outputs ``; inputs labelled. No forecast.ts create in this unattended watchlist run (per skill rules).

Anchors: FY2025 actual revenue ¥39.35B, net income ¥5.52B, EPS ¥7.64. FY2026 street consensus revenue ¥49.9–50.7B, EPS revised to ¥10.34 (down from ¥13.25 — analysts cutting on margin). A separate aggregator (Choyang, Jan 2026) put FY2026 revenue ¥46.79–52.02B and net ¥7.75–10.79B — a wide, aggressive net range I treat as upper-bound optimism.

Base-case forecast (would-be Brier log, not created this run): "002371.SZ FY2026 net income ≥ ¥9.0B, p≈0.40, resolves 2027-04." I put sub-coin-flip odds on ¥9B+ net — that requires margin recovery the FY2025/Q1-2026 trend argues against. Base EPS ¥10.3 vs current price ¥816–860 = forward P/E ~79–83x — priced for years of flawless execution.

Lens 12 · Bull vs Bear

Bull case. NAURA is the single best-positioned beneficiary of a structural, policy-mandated demand wave. Chinese fabs must hit ~70% local tooling by 2027; NAURA has the broadest domestic portfolio, the deepest installed base, state capital, and a backlog already booked into Q1-2027. Domestic WFE self-sufficiency is early — the China cohort is still just 6.5% of the global $41.4B WFE market, so the runway to import-substitute is enormous. Every US tightening adds to NAURA's addressable demand. Bundling breadth + consolidation (Kingsemi, Chengdu Guotai) widens the moat. If margins merely hold and revenue compounds 25%+, the earnings base doubles by FY2028.

Bear case (2–3 permanent-impairment risks).

1. Margin is structurally competed away, not cyclically dipped. The entire domestic cohort is losing gross margin while growing (NAURA 43.8%→39.5%, AMEC 45.8%→39.2%, Piotech 49.3%→33.3%). Government-orchestrated consolidation still leaves multiple state-backed rivals per process step with soft budget constraints — a recipe for permanent price competition and mid-30s% gross margins. A WFE vendor that can't hold margin is worth a fraction of ~100x earnings.
2. The valuation itself is the risk. ~107x trailing / ~80x forward, above ASML/AMAT/KLA, with the stock above the mean analyst target (¥560). Any growth or margin disappointment de-rates violently from this level; the A-share scarcity premium is not underwritable offshore.
3. Input-side sanction escalation. Presumption-of-denial already applies; a tightening to sub-component or software (EDA-for-tools, control firmware) could stall the very R&D breakthroughs (litho/implant/metrology) the bull case needs.

Pre-mortem (18 months out, thesis broke): it's early 2028. Revenue kept growing but gross margin fell through 36% as SMIC/CXMT played NAURA, AMEC, Piotech and a Kingsemi-adjacent field against each other on every tool; net income was flat for a third straight year; the street's ¥13–16 EPS never materialized; and the stock de-rated from 80x to 35x forward — a ~55% drawdown even with the business "working."

Contrarian view (what the market refuses to see): the market treats NAURA as a growth-and-share story and is right about the volume — but it is mispricing the margin. The Entity List guaranteed NAURA the revenue and simultaneously guaranteed it competition (Beijing won't allow single-vendor dependence; it funds rivals for resilience). The result is a high-volume, structurally-mid-margin utility wearing a hyper-growth multiple. The bull and bear are both right — just about different lines of the income statement.

Lens 13 · Devil's Advocate (short-seller)

Dismantling the bull case:

• Where's the revenue concentrated? In a handful of state-directed Chinese fabs (SMIC, CXMT, Hua Hong, YMTC) buying under mandate. If the China fab capex cycle cools (DRAM glut, logic overbuild, a domestic-demand air-pocket), NAURA's captive base becomes a captive concentration risk — and Beijing can throttle capex overnight.
• Why is the moat weaker than bulls think? Because Beijing's own policy is anti-moat. Industrial-security doctrine wants multiple domestic suppliers per process step, so the state actively funds NAURA's competitors (Piotech, AMEC, Hwatsing, ACM, SiCarrier and others). "China's AMAT" is the bull framing; "one of five state-funded vendors fighting on price" is the short framing — and the margin data (whole-cohort GM decline) already favors the short framing.
• Most dangerous competitor bulls underestimate: the emergent consolidators/entrants (e.g. Huawei-adjacent SiCarrier) with equal state backing and no need to earn a return — plus single-step specialists undercutting NAURA on their one tool.
• Worst capital-allocation / governance: dilutive state placements; M&A at policy prices; subsidy-dependent earnings; near-zero management equity alignment.
• What must hold for today's price? Sustained 25%+ revenue growth and margin recovery and successful litho/implant/metrology breakthroughs and no sanction escalation and the A-share scarcity premium persisting. That's a long conjunction at 80x forward.
• Growth disappoints 20–30%: at ~80x forward, a cut to FY2027 EPS of 25% plus a de-rate to a "mid-30s% margin utility" multiple (say 30–35x) implies a 50–65% price decline.
• Single permanent-impairment scenario: structural mid-30s% gross margins become the new normal (most plausible risk) → NAURA is a good ¥50B+ revenue business worth far less than ¥590B market cap.

Lens 14 · Management Questions (ordered by information value)

1. What was FY2025 gross margin by segment (semi-equipment vs precision components vs vacuum), and is the semi-equipment tool margin itself declining, or is the blended fall purely mix?
2. How much of FY2024 and FY2025 net income was government subsidies / grants booked through other income, and what is the run-rate ex-subsidy?
3. In competitive tool bids at SMIC/CXMT, how many domestic vendors do you now face per process step, and how has your win-price trended over the last four quarters?
4. What are FY2026–2027 contract liabilities (customer advances) and backlog by tool family, and how firm are the Q1-2027 backlog slots against cancellation?
5. What US-origin subsystems/components remain single-sourced post-Entity-List, and what share of BOM is not yet localized? What's the stockpile runway?
6. What is the realistic timeline and capex to a domestically credible ion-implant, metrology/inspection, and advanced-litho offering — build, buy (Kingsemi-style), or partner?
7. Where is the Kingsemi integration on synergies, and will you move to full consolidation/control? Same question for Chengdu Guotai.
8. What through-cycle operating-margin target are you underwriting, and at what revenue scale does R&D intensity fall below 12%?
9. How exposed is NAURA to a China fab-capex downturn (DRAM/logic overbuild), and what's your plan if domestic capex falls 20% in a year?
10. What is the aftermarket/service revenue mix today and the target — how much recurring margin can offset tool-price competition?
11. How do you weigh minority-shareholder returns vs state industrial objectives when they diverge (e.g. accepting low-margin national-priority orders)?
12. What is your capitalized-vs-expensed R&D policy, and would you commit to expensing to keep earnings quality clean?
13. What's the plan for the electronic-components line — divest, grow, or run as a strategic-materials hedge given its margin drag?
14. Any inbound/outbound M&A in litho, implant, or metrology under evaluation, and how would you fund it without further dilution?
15. If US controls tighten to tool-control firmware / design software, what is the contingency for your leading-edge (sub-14nm) roadmap?

---