ACM Research, Inc.

---

step: "01" title: Business Model & Overview ticker: ACMR company: ACM Research, Inc. source: coverage-next-full date: 2026-06-03

Step 01 — Business Model & Overview: ACM Research, Inc. (ACMR)

1. Business Description

ACM Research, Inc. (Nasdaq: ACMR) designs, develops, manufactures, and sells semiconductor process equipment used by integrated circuit manufacturers. [S1] The company's core competency is single-wafer wet processing — the removal of contaminants, residues, and particles from semiconductor wafers during fabrication. Its tools are used at multiple critical points in the chip manufacturing process, from front-end-of-line (FEOL) logic and memory fabrication to advanced packaging. [S1][S2]

ACMR was founded in 1998 and IPO'd on Nasdaq in November 2017. The company operates primarily through ACM Research (Shanghai), Inc. ("ACM Shanghai"), a majority-owned subsidiary incorporated in China and publicly listed on the Shanghai STAR Market since 2019. [S1] This dual-listing structure creates a complex governance and capital structure but provides deep local market relationships, RMB access, and Chinese government alignment. [S3]

2. Value Chain Positioning

ACMR sits in the equipment vendor tier of the semiconductor value chain:

Raw materials / chemicals
         ↓
Semiconductor equipment OEMs  ← ACMR operates here
(AMAT, LRCX, TEL, SCREEN, KLAC, ACMR)
         ↓
Fab customers (IDMs + foundries)
(SMIC, HLMC, CXMT, SK Hynix, Intel)
         ↓
Fabless design / chip customers
(Qualcomm, Apple, Nvidia, etc.)
         ↓
OEMs / End consumers

ACMR's position: A specialized second-tier equipment OEM with dominant share in single-wafer wet cleaning within China, growing into electrochemical plating (ECP) and advanced packaging equipment. Not yet a full-suite fab equipment provider — competes on specific tool categories rather than full-fab outfitting. [S2][S7]

3. Product Portfolio

Product Family (the "Planetary Family" — branded April 2026): [S3]

Product Category	Tools	Primary Use
Single-wafer wet clean	Ultra C (SAPS, TEBO, Tahoe)	FEOL particle/residue removal
Furnace/batch clean	—	Legacy batch processes
ECP (Electrochemical Plating)	Ultra ECP	Copper interconnect deposition
Advanced packaging	Ultra Packaging	Wafer-level packaging
Vertical furnaces	—	Thermal processing
PECVD	Saturn	CVD thin film deposition
Track coater/developer	—	Photoresist processing
Supercritical CO2 clean	—	Next-gen cleaning

Flagship: The Ultra C Tahoe single-wafer wet clean tool is the company's highest-margin, most differentiated product. It achieves 75% chemical reduction versus prior generation and controls particles down to 26nm. [S3] This is the key competitive differentiator versus SCREEN Holdings and legacy Western OEMs.

Revenue mix (estimated FY2025): [S2][S3]

• Single-wafer clean: ~65–70% of revenue
• ECP: ~15–20%
• Furnace / thermal: ~5–10%
• PECVD / Track / other: ~5–10% (Note: ACMR does not disclose exact product-level revenue split publicly)

4. Customer & Geographic Profile

Geographic concentration: [S1]

• Mainland China: ~99% of FY2025 revenue ($901M)
• International (SK Hynix, trial evaluations): <1%
• Oregon production facility under development for US-based customers (H2 2026 expected)

Top customers (estimated FY2025): [S7]

• HLMC (Hua Hong Microelectronics): ~15%
• SMIC (Semiconductor Manufacturing International): ~14%
• CXMT / memory fab successors: ~12% (YMTC entity successor context)
• Other China fabs: ~60%

Customer concentration: Top 3 customers ~41% of revenue; top 4 customers ~59% of receivables. [S7]

5. Revenue Model

Revenue recognition: Equipment revenue recognized upon delivery/acceptance of tools at the customer's fab. Service/support revenue (spare parts, maintenance) is smaller and recognized as delivered. [S1]

Ordering pattern: Multi-quarter lead times are typical. Customers place purchase orders 6–18 months in advance of expected tool delivery. This creates a backlog that provides near-term revenue visibility. [S3]

Pricing: Tool ASPs are not publicly disclosed. Based on comparable equipment in the sector, single-wafer clean tools typically price at $1M–$5M+ per unit depending on capability/node. ACMR competes on price versus SCREEN Holdings' higher ASPs. [S7]

6. Governance & Capital Structure

Dual-class share structure: [S5]

• Class A shares (NYSE): 1 vote/share — held by public investors
• Class B shares: 10 votes/share — held primarily by CEO David Wang
• CEO Wang controls 57.2% of voting power with only 14.4% economic interest

ACM Shanghai (STAR Market): ACMR holds ~70% of ACM Shanghai equity. The September 2025 private placement raised RMB 4.5B ($623M) at ACM Shanghai level, diluting ACMR's stake slightly and creating a $466M Non-Controlling Interest on the consolidated balance sheet. [S2] This capital is trapped at the Shanghai subsidiary level (subject to Chinese capital controls and SAFE regulations).

Summary balance sheet (FY2025 year-end): [S1][S2]

• Cash & equivalents: $1.17B (consolidated); Net cash attributable to ACMR: $845.5M
• Short-term borrowings: ~$120M
• Long-term debt: minimal
• NCI: $466M
• Total equity (ACMR-attributable): ~$1.05B

7. Strategic Positioning

Core thesis: ACMR is the beneficiary of China's semiconductor self-sufficiency policy (the "localization" mandate). Western equipment OEMs face escalating export controls under the BIS rules of Oct 2022, Oct 2023, and the December 2024 Entity List designation of ACM Shanghai. [S6] Meanwhile, China fabs (SMIC, HLMC, CXMT) have been directed to localize supply chains — creating an effectively captive market for ACMR within China. [S7]

International pivot: Management has set a long-term aspiration of generating "half revenue outside China." The most tangible proof point is the SK Hynix ECP/HBM order (>RMB 200M, Q3 2025). The Oregon manufacturing facility (opening H2 2026) is designed to serve US-based customers without involving ACM Shanghai (the Entity-Listed subsidiary). [S3]

Key tension: The very geopolitical dynamic that created ACMR's captive China market also limits its global addressability. Tools manufactured through ACM Shanghai cannot be easily exported to US allies. The Oregon facility is a structural workaround, but it starts small and faces a long ramp. [S3][S6]

---

Source Index

• [S1] SEC EDGAR XBRL + 10-K filings for ACMR (CIK 1680062)
• [S2] StockAnalysis.com financial statements and statistics
• [S3] ACM Research FY2025 10-K (filed 2026-03-02); Q4/FY2024 earnings press release
• [S4] Street consensus via Tavily WebSearch (June 2026)
• [S5] ACM Research DEF 14A proxy (filed April 29, 2025)
• [S6] BIS Entity List Federal Register notices (Dec 2024); BIS Oct 2022 / Oct 2023 rules
• [S7] Competitive intelligence: SCREEN Holdings, NAURA, Kerrisdale Capital thesis (Jan 2025)

---

step: "04" title: Financial Quality & Adversarial Sweep ticker: ACMR company: ACM Research, Inc. source: coverage-next-full date: 2026-06-03

Step 04 — Financial Quality & Adversarial Sweep: ACM Research, Inc. (ACMR)

1. Statement Quality Assessment

Revenue Recognition Quality: MODERATE-HIGH

ACMR recognizes revenue upon customer acceptance of tool deliveries. [S1] The business model (discrete equipment sales to identifiable customers under purchase orders) is relatively straightforward. Revenue cannot be easily manipulated through channel stuffing because fab customers are sophisticated, the verification process is rigorous, and the SEC EDGAR filing history shows consistent year-end reconciliations.

Watch point: China-based customers may have longer acceptance processes; some revenue recognition depends on customer sign-offs that could be delayed or front-loaded quarter-to-quarter. Quarterly variability is high (Q4 tends to be strongest, Q1 weakest). This is sector-normal and not a red flag. [S1]

Earnings Quality: MODERATE

Metric	FY2023	FY2024	FY2025
Net Income (ACMR attr.)	$74.3M	$103.6M	$94.1M
Operating Cash Flow	($49M)	$141M	~($10M)
CFO / Net Income Ratio	(0.66x)	1.36x	(~0.11x)

[S1][S2] The FY2023 and FY2025 negative operating cash flows despite positive GAAP earnings are a concern. The culprit is working capital (inventory build) rather than earnings manipulation. FY2024 was the first year of genuine cash conversion ($141M OCF vs. $103.6M net income). FY2025 returned to negative OCF due to inventory pre-purchasing (Entity List response). [S3]

Assessment: Earnings are real but cash conversion is poor and lumpy. Not a fraud signal — this is structural to the business model. However, it means GAAP earnings overstate economic earnings in most years.

Balance Sheet Quality: MODERATE (complex)

The consolidation of ACM Shanghai creates complexity:

• $1.17B cash on consolidated balance sheet includes funds trapped at the Shanghai subsidiary level, subject to Chinese capital controls
• $845.5M net cash attributable to ACMR is the cleaner metric [S3]
• $466M NCI (non-controlling interest) from ACMSH September 2025 equity raise — dilutes economic ownership of ~$623M Shanghai cash
• $703M inventory is real but carries risk of obsolescence if customer delivery schedules slip

Adjustments to GAAP

Item	Direction	FY2025 Est. Impact
Stock-based compensation	Add back	+$35–45M
NCI adjustment	Remove from economic earnings	($27M)
Inventory provision (one-time)	Add back	+$10–20M
Component re-sourcing premium	Ongoing (no add-back)	Structural
Adjusted Net Income (est.)	—	~$115–130M

2. Adversarial Research Sweep

Note: Transcript analysis was not performed (coverage-next-full path). Bear arguments sourced from published short theses, analyst reports, SEC disclosures, and press coverage.

Bear Thesis #1 — Kerrisdale Capital (January 2025): "Pattern Collapse at FinFET Nodes"

Source: Kerrisdale Capital published a long-form bear thesis on ACMR in January 2025. [S7]

Core claim: Kerrisdale argued that ACMR's cleaning technology is optimized for 28nm and older nodes. As China fabs attempt to push toward 7nm/5nm FinFET, the pattern density requires different cleaning physics — specifically, pattern collapse becomes a challenge for liquid-based wet cleaning. Kerrisdale suggested ACMR's tools may have a natural ceiling in advanced node capabilities.

Counter-evidence: The Ultra C Tahoe explicitly addresses <26nm particle control, and ACMR has announced supercritical CO2 clean tools for sub-10nm applications. The FY2026 product launch pipeline (Saturn PECVD, supercritical CO2) suggests R&D is ahead of the node transition. However, Kerrisdale's core point — that international fabs at 3nm/2nm nodes may not be addressable with ACMR tools — has not been fully rebutted. [S3]

Verdict: Partially valid for international expansion ambitions; less relevant for China-node market (28nm–14nm where ACMR is proven). Flag as a medium-term bear risk. [A01]

Bear Thesis #2 — BIS Entity List Escalation Risk

Source: Dec 2024 Federal Register; company 10-K risk factors. [S3][S6]

Core claim: ACM Shanghai was added to the Entity List in December 2024. The bear case argues: (1) the supply chain disruption is more severe than management admits, (2) the parent company ACMR could be added in subsequent rounds, and (3) Chinese customers may preemptively reduce ACMR tool purchases given the US government targeting signal.

Counter-evidence: FY2026 guidance reaffirmed at $1.08B–$1.18B. Management stated supply chain reorganization is "manageable." Q1 2026 shipment decline ($157M vs $245M prior year) appears partially Entity-List-driven, but H2 recovery expected based on backlog. [S3]

Verdict: The escalation tail risk is real and not fully priced. If ACMR parent (not just ACM Shanghai) is ever Entity Listed, the business faces an existential threat. Probability: low in the near term but non-zero given geopolitical direction. [A02]

Bear Thesis #3 — China Customer Credit Risk

Source: Kerrisdale Capital; sector analyst commentary. [S7]

Core claim: Some Chinese fab customers are operating with state support rather than commercial viability. If China's semiconductor buildout faces policy reversal or economic stress, customer payment risk increases. ACMR has ~$700M+ inventory partially destined for customers who might delay acceptance.

Counter-evidence: SMIC and HLMC are publicly listed companies with audited financials. CXMT (YMTC successor) is more opaque. ACMR does take credit risk on tools in work-in-progress, but the payment terms are largely secured by advance deposits and letters of credit for large orders.

Verdict: Moderate concern. Not an immediate catalyst but a tail risk in a Chinese economic stress scenario.

Bear Thesis #4 — Dual-Class Governance Discount

Source: DEF 14A; governance analysis. [S5]

Core claim: CEO David Wang controls 57.2% of votes with 14.4% economic stake. Minority shareholders have no effective governance recourse. Wang has been systematically selling shares ($59.9M over 24 months) via 10b5-1 plans, despite controlling the company.

Counter-evidence: Dual-class is common in tech/founder-led companies. Wang built the company from scratch and his interests are partially aligned (14.4% economic stake is still meaningful). The selling is programmatic and disclosed.

Verdict: Real governance risk but par for the course in founder-controlled tech companies. Represents a structural valuation discount vs. peers with clean governance.

No Fraud Allegations or SEC Investigation Found

A search of SEC enforcement actions, class action securities litigation, and short-seller reports found no allegations of revenue fabrication, undisclosed related-party transactions, or PCAOB audit failures for ACMR. [S1][S7] The HFCAA (Holding Foreign Companies Accountable Act) risk was referenced in prior years but PCAOB inspections of ACM Shanghai's auditor appear to have been completed satisfactorily.

3. Summary: Financial Quality Rating

Dimension	Rating	Notes
Revenue recognition	B+	Clean method; quarterly lumpiness normal
Earnings quality	B-	Poor cash conversion most years; working capital drag
Balance sheet transparency	B	Complex NCI + trapped China cash; no fraud signal
Governance	C+	Dual-class structure; systematic insider selling
Audit quality	B	Big-4 auditor; PCAOB compliant
Overall	B	Real business, real revenue; financial complexity warrants discount

---

Source Index

• [S1] SEC EDGAR XBRL + 10-K filings (CIK 1680062)
• [S2] StockAnalysis.com financial statements
• [S3] ACM Research FY2025 10-K (filed 2026-03-02)
• [S4] Street consensus via Tavily WebSearch
• [S5] DEF 14A proxy (April 2025)
• [S6] BIS Federal Register Entity List notices (Dec 2024)
• [S7] Kerrisdale Capital ACMR bear thesis (January 2025); competitive intelligence

---

step: "12" title: Bull/Bear Debate & Catalysts ticker: ACMR company: ACM Research, Inc. source: coverage-next-full date: 2026-06-03

Step 12 — Bull/Bear Debate & Catalysts: ACM Research, Inc. (ACMR)

Note: Transcript analysis was not performed (coverage-next-full path). The bull/bear debate is inferred from consensus reports, press releases, the Kerrisdale Capital short thesis (Jan 2025), SEC filings, and industry analysis.

1. The Central Debate

The core disagreement about ACMR is whether the company is:

Bull framing: A high-quality Chinese equipment champion with a captive domestic market, a believable international expansion story, and a balance sheet backstop from the ACMSH equity raise — temporarily disrupted by the Entity List but fundamentally on track for $1B+ revenue in FY2026 and continued growth.

Bear framing: A single-country equipment story with a regulatory tailwind that is also its ceiling — trapped inside China's borders by the same geopolitical dynamics that gave it market share, facing structural margin compression, and with international expansion ambitions that are unproven at any material scale.

2. Bull Case Arguments

Bull 1: China Self-Sufficiency Is a Multi-Decade Policy Commitment

China's semiconductor localization policy is not a short-term initiative. The government has invested hundreds of billions in fab construction and the CICSF "big fund" has multiple rounds of committed capital. SMIC, HLMC, and CXMT have multi-year capex budgets that require domestic equipment suppliers to fill. ACMR is the only domestic supplier with proven single-wafer wet clean tools at mature nodes. [S6][S7]

Even with Entity List disruption, FY2026 guidance was reaffirmed at $1.08B–$1.18B — a 20–30% growth year. China capex commitments are largely pre-placed in purchase orders, providing 12–18 months of backlog visibility. [S4]

Bull 2: Entity List Creates Opportunity, Not Just Risk

A counterintuitive bull argument: by designating ACM Shanghai, the US government has signaled to China fabs that ACMR is a "national champion" — the kind of company that China will double down on supporting, qualifying, and purchasing from. The Entity List designation may accelerate Chinese customer loyalty and state support (subsidies, fast-tracking qualifications). [S3][S7]

Bull 3: International Credibility Is Inflecting

The SK Hynix ECP win for HBM applications (>RMB 200M, Q3 2025) is the first genuine proof point that ACMR tools can pass qualification at a leading non-Chinese fab. HBM is the hottest category in semiconductor equipment (driven by AI/Nvidia demand). If ACMR can win follow-on orders from SK Hynix or Micron, the "China-only" narrative breaks down. [S3][S7]

The Oregon manufacturing facility removes the Entity List barrier for Western customers. If operational by H2 2026, it enables ACMR to pitch ASML/TSMC-ecosystem tools to Intel, Samsung, and Micron without supply chain concerns. [S3]

Bull 4: Gross Margin Recovery Visible in the Data

FY2025 gross margin bottomed at 41.3% in Q1 2025 and has been recovering: Q2 2025 (43.7%) → Q3 2025 (45.4%) → Q4 2025 (46.1%). Management guided H2 2026 margins to the middle of the 42%–48% range. If new high-margin products (SPM clean, supercritical CO2) ramp in H2 2026, 47–48% gross margin is achievable. Earnings leverage from margin recovery on a $1.1B+ revenue base would be substantial. [S3]

Bull 5: Net Cash Optionality

The $845.5M ACMR-attributable net cash position is 13% of market cap. This provides:

• Buffer against negative FCF cycles
• Optionality for accelerating Oregon facility
• Potential for strategic acquisitions in international markets
• Downside protection in a valuation reset scenario [S1][S2]

3. Bear Case Arguments

Bear 1: Pattern Collapse at Advanced Nodes Limits International TAM

Kerrisdale Capital's core technical argument: ACMR's wet clean tools are optimized for 28nm and above, where pattern density is manageable with liquid-based cleaning. At 5nm/3nm FinFET geometries, the surface tension of cleaning liquids causes pattern collapse — wafer features bend and break during cleaning. SCREEN Holdings has addressed this with specialized tools; ACMR's supercritical CO2 tool is still in development. [S7]

Implication: ACMR cannot address the global leading-edge market (TSMC, Samsung Foundry, Intel 18A) until its next-gen tool is proven. The SK Hynix ECP win is for plating, not cleaning — it does not rebut this thesis. International addressability may be limited to mature nodes for several more years.

Bear 2: Entity List Supply Chain Cost Is Structural, Not Transitory

The gross margin compression from 50%+ to 44% is presented as partially transitory (one-time inventory provisions) but partly structural. Component re-sourcing from non-US vendors permanently increases Bill of Materials costs. ACMR estimated this at ~150bps drag. If the Entity List is not reversed (high probability), this drag persists indefinitely. [S3][S7]

In addition, ACMR must now operate two manufacturing systems (Shanghai-based for China; Oregon-based for international). Dual manufacturing increases fixed costs and dilutes operating leverage.

Bear 3: Customer Concentration + Receivables Risk

The top 3 customers (HLMC ~15%, SMIC ~14%, CXMT ~12%) represent ~41% of revenue. [S7] These are large state-owned enterprises with:

• Political buying decisions that can reverse on government direction
• Payment terms that extend working capital cycles
• The ability to demand price reductions given ACMR's dependence on them The $420M accounts receivable balance with 59% concentrated in 4 customers represents meaningful credit risk if China's economy faces stress. [S1]

Bear 4: International Expansion Timeline Is Optimistic

Management has stated a long-term goal of "half revenue outside China." At <1% today, this requires 50x growth in international revenue — an extraordinary ambition given:

• SCREEN, AMAT, and LRCX dominate non-China customers with 20+ year relationships
• Oregon facility starts at very small scale
• ACMR has limited brand recognition outside China Even the most bullish international scenario (10–15% of revenue by FY2028) requires multi-year execution with no guarantee of success. [S3][S7]

Bear 5: Dual-Class Structure Creates Permanent Discount

CEO Wang controls 57.2% of votes. Minority shareholders cannot effectively influence capital allocation, compensation, or strategic direction. This governance structure permanently discounts ACMR's valuation vs. peers with one-share-one-vote. [S5] The risk is that Wang optimizes for different objectives (e.g., China government relations, ACMSH listing value) vs. Nasdaq shareholder value.

4. Catalyst Map

Positive Catalysts (Bull)

1. Q2–Q4 2026 revenue recovery: If FY2026 first-half revenue bounces back from Q1 2026's $157M shipment trough → validates FY2026 guidance → stock re-rates upward
2. Gross margin recovery to 46–48%: Any quarter printing ≥47% GM sends a strong signal that Entity List cost drag is behind peak
3. Second non-China customer win: If ACMR announces a second international ECP or clean tool order (especially from a US or South Korean fab) — demonstrates the SK Hynix win was not a one-off
4. Oregon facility first tool delivery: Concrete evidence the international manufacturing strategy is real
5. US-China tech détente: Any diplomatic signal that BIS will not escalate further (unlikely in near term but would be a material re-rating catalyst)

Negative Catalysts (Bear)

1. ACMR parent added to Entity List: Existential — would crater the stock 50%+
2. FY2026 guidance cut: If revenue guidance is reduced from $1.08–$1.18B → signals Entity List impact is worse than disclosed
3. Gross margin fails to recover: Sustained GM below 43% in Q2–Q3 2026 would validate the structural degradation bear case
4. Major China customer loss: SMIC or HLMC materially reducing orders (either political or commercial)
5. Earnings restatement / audit issue: PCAOB / HFCAA escalation with ACM Shanghai auditor

---

Bull Case — 3 Bullets

• China's semiconductor localization policy is a multi-decade commitment; ACMR is the only proven domestic single-wafer wet clean supplier, making it a national champion with a captive $800M+ annual addressable market in China alone
• Gross margin recovery from Q1 2025's 41.3% trough is visible quarter-over-quarter (46.1% in Q4 2025); management's 42–48% long-term range is achievable, creating earnings leverage on $1.1B+ FY2026 revenue
• The SK Hynix HBM ECP win and the Oregon manufacturing facility together represent the first credible evidence that ACMR's "half revenue outside China" vision is a real strategy, not just a talking point — optionality that current valuation (70x PE) only partially reflects

Bear Case — 3 Bullets

• Entity List designation for ACM Shanghai has permanently elevated manufacturing costs by ~150bps of gross margin and created dual-system manufacturing complexity that dilutes future operating leverage — structural, not transitory
• ~99% China revenue concentration exposes shareholders to both existential tail risk (ACMR parent Entity Listing) and a one-directional customer concentration dynamic where SMIC/HLMC hold negotiating power that erodes pricing and margins over time
• International expansion is a multi-decade aspiration, not a near-term growth driver — at <1% of revenue today and facing entrenched SCREEN/AMAT dominance globally, ACMR will remain a China-captive equipment supplier for at least 3–5 years, putting a structural ceiling on its valuation multiple

---

Source Index

• [S1] SEC EDGAR XBRL (CIK 1680062)
• [S2] StockAnalysis.com
• [S3] ACM Research FY2025 10-K; press releases; FY2026 guidance
• [S4] Street consensus estimates (June 2026)
• [S5] DEF 14A proxy (April 2025)
• [S6] BIS export control; China semiconductor policy; SEMI.org
• [S7] Kerrisdale Capital bear thesis (Jan 2025); competitive intelligence; customer data