China ADRs at the Qwen 3.7-Max moment: 22 names across 5 layers, 3 portfolios — June 2026 guide for Indian investors

Chinese ADRs are an unusual entry on the 2026 thematic roster. They are simultaneously the highest-intent SEO category — every Indian retail investor with US-brokerage access knows BABA, PDD, JD by name — and the most structurally loaded category of any thematic guide. This article holds both truths.

Six events have rewritten the China-ADR setup in twelve months:

January 20, 2025 (DeepSeek R1): DeepSeek R1 triggered an 18% one-day drop in Nvidia and a global re-rating of China AI capabilities. Since then: DeepSeek V3-0324 (March 2025), R1-0528 (May 2025), V3.1 (August 2025), V3.2 (December 2025), V4 (April 2026). Alibaba's Qwen3.7-Max (May 19, 2026): #5 globally, #1 Chinese on Artificial Analysis Intelligence Index (56.6), ahead of Google Gemini 3.5 Flash. Baidu's ERNIE 5.1 (May 8, 2026): #1 Chinese / #4 global on LMArena Search Leaderboard at ~6% of comparable-model pre-training cost.

February 20, 2026 (SCOTUS IEEPA ruling): Supreme Court ruled the President cannot use IEEPA to impose tariffs — restructuring much of the broad authority basis. Section 301 tariffs untouched.

November 2025 (Trump tariff truce extended): Truce extended through November 10, 2026, lowering fentanyl tariff to 10% (had been 20%), keeping reciprocal tariff at 10%. Current effective US tariff rate on Chinese imports ~39% average (10% reciprocal + 10% fentanyl + Section 301 stacks).

December 22, 2025 (ZEEKR delisted from NYSE): Geely took ZK private at $26.87/ADS. The take-private path of "delisting" — distinct from HFCAA enforcement and DiDi-style cybersecurity-review delisting. Indian investors should know all three pathways exist.

May 27, 2025 (BeOne Swiss redomicile): BeiGene renamed to BeOne Medicines, redomiciled from Cayman Islands to Basel Switzerland (Cayman → Swiss "continuation" process), maintains triple listing NASDAQ ONC + HKEX 06160 + SSE 688235. The cleanest structural de-risking of any Chinese ADR.

Q1 2026 prints (May 2026): BABA Q4 FY26 cloud +38% YoY; Qwen ranked #5 global. PDD Q1 missed consensus on EPS by RMB 7.26 — stock fell -8 to -11%. Net cash ~$60 billion (~half of market cap). JD revenue +4.9%; food delivery 10% share burning ~$35B in subsidies last 3 quarters of 2025. BIDU AI Cloud +79% YoY; GPU Cloud +184% YoY — the cleanest training-demand signal in the China stack. Apollo Go 3.2 million driverless rides Q1, +120%+ YoY; cumulative 22M+ rides, 190M+ driverless kilometers.

This guide is for the Indian-resident investor who wants real exposure to China's AI re-rating but isn't pretending the delisting + tariff + VIE structure risks aren't structural.

What this guide is and isn't

It is: 22 names organized by what the China AI thesis, the HFCAA risk, and the tariff state actually reward, one verdict per name, three model portfolios — including an explicit "0% is defensible" option.

It is not: "China is cheap, buy it." Not pretending VIE structure or delisting risk has been resolved. Not assuming Trump tariffs stay at current levels indefinitely.

A note on data. Every revenue, AI benchmark, capex figure, and regulatory disclosure traces to a primary source — SEC 20-F/6-K, CMS/Treasury documents, PCAOB filings, China company IR pages.

The framework: dual structural truth

The bull case is China AI. BABA Cloud Intelligence Group Q4 FY26 revenue RMB 41.6 billion (+38% YoY); AI-related revenue triple-digit YoY for 11 consecutive quarters; exceeded 30% of external cloud commercialization. CEO Eddie Wu guided AI to >50% of cloud segment within 12 months. BIDU GPU Cloud +184% YoY. Qwen3.7-Max + ERNIE 5.1 + DeepSeek V4 are the three credible Chinese frontier models.

The bear case is structural delisting + VIE + tariff risk. HFCAA is "two-bad-years-in-a-row away from delisting" — politically driven, outside investor control. House Select Committee on CCP sent multiple letters in 2025 demanding SEC delist CCP-linked companies — naming Alibaba, Baidu, Hesai, Zeekr explicitly. SEC Chair Atkins (confirmed April 2025) publicly stated his agency will increase enforcement. Trump Feb 21, 2025 "America First Investment Policy" memo flagged VIEs and Chinese audit oversight for review.

The third truth: most major ADRs are now HK dual-primary. BABA upgraded August 2024; JD, BIDU, NTES, BILI dual-listed (most secondary, several upgraded). Practical implication: Indian retail investor whose US broker holds Chinese ADR has theoretical conversion access to HK shares — but conversion mechanics, fees, and broker support vary widely (IBKR best HK access; Vested/INDmoney/Rovia US-focused).

Five layers:

Layer	What it is	2026 read
1. Mega-cap China internet	BABA, PDD, JD, BIDU, NTES	AI re-rating thesis lives here
2. Consumer + travel	TCOM, YUMC, BILI, VIPS, WB, HTHT	Consumption softness but stable franchises
3. China EV	NIO, XPEV, LI, BYDDY	(see automotive guide)
4. China fintech	FUTU, TIGR	Both took CSRC penalties Q1 2026
5. China biotech	ONC (formerly BeOne/BeiGene)	Swiss redomicile = cleanest structural de-risk

Verdict format:

Verdict — [Action]: [Reason]. [Caveat].

---

Layer 1 — Mega-cap China internet

BABA — Alibaba Group

Q4 FY26 (CQ1 2026, reported May 13, 2026): total revenue RMB 243.38B (~$35.3B), +3% YoY. Beat on cloud + AI metrics the market actually trades on.

Cloud Intelligence Group: RMB 41.6B (~$6.04B), +38% YoY. External commercial cloud revenue accelerated to +40% YoY. AI-related product revenue triple-digit YoY for 11 consecutive quarters, contributing RMB 9B ($1.3B) to cloud quarter and exceeding 30% of external cloud commercialization for first time. CEO Eddie Wu guided AI to >50% of cloud segment within 12 months.

TTG (China commerce): CMR +8% YoY like-for-like. Quick commerce ("Taobao Instant Commerce") revenue +57% YoY to RMB 19.9B — aggressive subsidy war vs Meituan + JD. China e-commerce Adjusted EBITA fell ~40% YoY on instant-commerce investment.

Group adjusted EBITA down 84% YoY to $740M; adjusted net profit fell 99.7% to RMB 86M ($12.7M); FCF swung to -$2.51B. The collapse is deliberate: AI capex + quick-commerce subsidies.

Net cash ~$41 billion at March 31, 2026 (down from ~$50B peak as buyback + capex draw down). Buyback $19.1B remaining under authorization through March 2027. Capex RMB 380B (~$53B) over three years FY26-FY28 announced February 2025; Q4 FY26 AI+cloud capex hit record RMB 38.6B single quarter.

Qwen3.7-Max (released May 19, 2026): Artificial Analysis Intelligence Index 56.6 — 5th overall, #1 Chinese, ahead of Google Gemini 3.5 Flash. GPQA Diamond 92.4 (vs Claude Opus 4.6 Max 91.3). HMMT Feb 2026: 97.1 (highest in table). SWE-bench Verified 80.4. 1M token context window. 35-hour autonomous execution demo (1,158 tool calls). API pricing $2.50/M input, $7.50/M output. Supports external harnesses including Anthropic's Claude Code.

HK dual-primary listing completed August 2024 (HKEX 9988). Eligible for Stock Connect.

Forward P/E ~19× FY27e — but sum-of-parts (cloud + AI valued separately + cash) implies substantial upside if cloud growth holds.

Catalyst: Q1 FY27 print (August 2026) — clarity on whether profit drawdown is 1-2 quarter phenomenon or longer; Qwen adoption metrics.

Risk: Profit collapse will read as competitive necessity to some investors; VIE + HFCAA overhang structural; cloud growth could decelerate as easy comps roll off.

Verdict — Add (advanced): The cleanest China AI thesis at the lowest multiple. Qwen3.7-Max validates the platform. Cash + buyback is the floor. Size 1-3% of equity portfolio max given delisting risk.

PDD — PDD Holdings (Pinduoduo + Temu)

Q1 2026 (reported May 27, 2026): total revenue RMB 106.2B (~$15.4B), +11% YoY — missed consensus RMB 109.82B. Transaction services revenue (Temu proxy) RMB 56.3B (~$8.2B), +20% YoY. Online marketing services (Pinduoduo China advertising) RMB 49.9B (~$7.2B), +2.5% YoY — the historical cash cow decelerating to single digits.

Profitability: Adjusted net income RMB 14.1B (~$2.0B), -17% YoY. Adjusted EPS RMB 9.51 (~$1.38) missed consensus RMB 16.77 by RMB 7.26. Stock fell -8 to -11% on print.

Temu deceleration specifics:

• De minimis ended May 2, 2025 for China/HK; globally August 29, 2025; locked Feb 2026 EO
• De minimis effective rate ~54%
• Temu pivot to semi-managed + US-warehouse model. US profit margins reportedly improved from 2% to 4% under semi-managed model
• Nov 2025 90-day tariff truce (extended Nov 10, 2026) cut small-package rate by more than half from peak — but structural exemption is gone

Net cash position ~$60 billion as of Q1 2026 print — nearly half of market cap. The cash-rich-but-misunderstood story is the bull anchor.

Forward P/E ~7-9× FY27e — among cheapest mega-caps globally.

Risk: Pinduoduo China advertising decelerated to single digits; Temu structural model broken (cross-border-cheap-package that built it); $60B cash real but management unwilling to return (no dividend, no buyback announced).

Verdict — Hedge: $60B cash at half of market cap is the floor; Temu pivot execution is the upside; management capital allocation is the bear case. Size 1-2%.

JD — JD.com

Q1 2026 (reported May 13, 2026): total revenue RMB 315.7B (~$43.5B), +4.9% YoY — beat consensus.

Segment performance:

• JD Retail (core): general merchandise revenue accelerated to +14.9% YoY; advertising + commission "strong double-digit growth"
• JD Logistics: revenue RMB 60.58B (~$8.37B), +29% YoY; net income RMB 880M, +95.2% YoY
• New Businesses (incl. food delivery): significant sequential loss reduction

Food delivery (launched February 2025) — the competitive battleground:

• JD daily orders ~9-10M Q1 2026 vs Meituan ~65M (~55% share) and Alibaba's Taobao Flash Purchase ~50M (~35%). JD share ~10%.
• Aggregate three-player subsidies last 3 quarters of 2025: ~RMB 145B burned (Meituan ~40B + Alibaba ~70B + JD ~35B)
• Beijing summoned all three in 2025 to cool the subsidy war — order book stabilizing per Q1 2026 commentary

Net cash RMB 215.7B (~$30 billion) at March 31, 2026.

Forward P/E ~8-9× FY27e.

Risk: Structurally lower margin than BABA (1P heavy); food delivery is P&L drag with uncertain steady-state share above 10%; Pinduoduo's encroachment on JD's middle-class consumer is well-documented.

Verdict — Add: JD Logistics is structurally undervalued inside the parent; food delivery share at 10% is the leg-up bet. Cheap multiple compensates for execution risk.

BIDU — Baidu

Q1 2026 (reported May 21, 2026): total revenue RMB 32.08B (~$4.43B). General Business revenue RMB 26.0B, +2% YoY.

AI Cloud — CRITICAL bull-case driver:

• AI Cloud Infra revenue +79% YoY
• GPU Cloud revenue +184% YoY — the cleanest training-demand signal in China stack
• Core AI-powered business revenue RMB 13.6B (+49% YoY) — crossed 50% of Baidu General Business mix for first time

ERNIE Assistant: DAUs nearly doubled YoY in March 2026; daily conversation rounds more than tripled.

ERNIE 5.1 (released May 8, 2026):

• Compressed total parameters to ~1/3 vs ERNIE 5.0
• 6% of comparable-model pre-training cost
• LMArena Search Leaderboard: 1,223 points — #1 China / #4 global
• Agent capabilities surpass DeepSeek-V4-Pro

Apollo Go (autonomous robotaxi):

• Q1 2026: 3.2 million fully driverless operational rides, +120%+ YoY
• Weekly rides peaked at 350,000+ in March 2026
• Cumulative rides exceeded 22M by April 2026; 190 million fully-driverless kilometers
• International expansion: South Korea, Dubai ride-hailing app launched Q1 2026

Net cash position total cash + investments RMB 279.3B (~$40.49 billion) at March 31, 2026.

Forward P/E ~8-10× FY27e — sum-of-parts (Apollo Go alone toward $30B+ if compared with Waymo) makes search approach free-option territory.

Catalyst: Apollo Go quarterly ride growth; Dubai + South Korea volumes; ERNIE 5.1 enterprise adoption; AI Cloud growth durability.

Risk: Search ads continue secular decline; Apollo Go not yet meaningful revenue driver (millions of rides at low ASP); ERNIE 5.1 commercialization unclear; iQIYI structural drag.

Verdict — Add: Best risk-adjusted China AI bet — AI Cloud +79% + Apollo Go scaling + 40B+ net cash at ~9× FY27e. Size 1-2%.

NTES — NetEase

Q1 2026 (reported May 22, 2026): total revenue RMB 30.6B (~$4.4B) vs RMB 28.8B PY. Games + related VAS RMB 25.7B (~$3.7B, +6.9% YoY) = 84% of revenue. Gross profit margin 69.4% (vs 54.1% PY) — significant expansion. Net income attributable to shareholders RMB 10.7B ($1.5B).

Capital return: Q1 dividend $0.144/share ($0.720/ADS); Q4 2025 dividend $0.232/share ($1.16/ADS). $5.0B buyback program extended November 20, 2025 by 36 months through January 9, 2029. 23.2M ADSs repurchased for $2.1B cumulative.

HK secondary listing (HKEX 9999); dual-primary path candidate.

Forward P/E ~14-16× FY27e — the cleanest "boring + cash-returning + structural China gaming franchise"; lowest political risk of the mega-caps.

Verdict — Core buy: Best risk-adjusted China internet exposure. Gaming franchise quality + dividend + buyback + lowest political risk profile.

---

Layer 2 — China consumer + travel

TCOM — Trip.com Group

Q1 2026 (reported May 19, 2026): consensus revenue $2.33-2.35B, EPS $0.85-1.05 (specific Q1 RMB breakdowns verify via TCOM IR/6-K). FY 2025 context: net revenue RMB 62.4B (~$8.9B, +17% YoY); international OTA bookings +60% YoY; outbound bookings >120% of pre-COVID; international = 40% of revenue.

Forward P/E ~18-22× FY27e.

Verdict — Add: Travel franchise quality + outbound recovery + Trip.AI agentic search. Cleanest consumption-China play.

YUMC — Yum China

Q1 2026 (reported May 5, 2026): revenue $3.27B (+10% YoY). Operating profit $447M. Diluted EPS $0.87 (+7% YoY). KFC same-store sales +1% (4th consecutive positive quarter); restaurant margin 19.1%. Pizza Hut operating profit +18% YoY; entered 100+ new cities Q1 driven by WOW small-format. Store count 18,737 (KFC 13,454, Pizza Hut 4,375); 636 net new stores in Q1; on track to surpass 20,000 stores in 2026. Delivery ~54% of company sales.

Forward P/E ~18-20× FY27e.

Note: YUMC is not VIE-structured in the same way as internet ADRs (Cayman parent but operates restaurants — different ownership-structure profile).

Verdict — Core buy (defensive): Quality franchise + non-VIE structure + dividend + buyback. Best risk-adjusted China consumer exposure.

BILI — Bilibili

Q1 2026 (reported May 21, 2026): revenue RMB 7.47B (~$1.04B, +7% YoY). Advertising +30% YoY — 13th consecutive quarter double-digit growth (AI-improved ad targeting). VAS RMB 2.9B +4%. Mobile games RMB 1.52B (~$223M, -12% YoY). Net profit RMB 202M vs net loss PY. Adjusted net profit RMB 585M (~$86M, +62% YoY). Gross margin 37.1% — 15th consecutive quarter expansion. DAU 115M (+8%); MAU 376M.

Forward P/E ~25-30× FY27e.

Verdict — Add: Profitability inflection real; ads franchise compounding. Premium multiple supported by 15-quarter margin expansion streak.

VIPS, WB, HTHT

Vipshop (VIPS): Q1 revenue RMB 26.6B (+1.2%); net income +13.6% YoY. Q2 2026 guide -5% to flat (soft).

Weibo (WB): Q1 revenue $421.3M (+6%); advertising +9% YoY.

H World (HTHT): Q1 revenue CNY 6.0B (+11.1%); 13,215 hotels; HWC RevPAR CNY 214.

Verdict (all three) — Hold: Defensive consumption franchises; lower growth than mega-caps.

---

Layer 3 — China EV (reference only)

NIO, XPEV, LI, BYDDY — see automotive guide for full coverage.

ZK (ZEEKR) DELISTED from NYSE December 22, 2025 — Geely take-private at $26.87/ADS. Indian retail can no longer hold.

---

Layer 4 — China fintech (online brokerages)

FUTU — Futu Holdings

Q1 2026 (reported May 28, 2026): revenue HK$5,856M (~US$746.9M, +24.7% YoY) — missed Zacks consensus US$773M by 3.33%. Brokerage commissions HK$2,641.4M (~US$337M, +14.3% YoY). 30.17M registered users; 3.59M funded accounts; total client assets $155.8B (+47.2% YoY).

CRITICAL NEGATIVE: Net income per ADS HK$6.00 (~$0.77), -60.7% YoY — driven by RMB 1.85B (~$255M) penalty from CSRC Shenzhen Bureau for unlicensed cross-border brokerage activity, fully reflected in Q1.

Forward P/E ~12-15× FY27e ex-fine.

Risk: CSRC penalty signals ongoing China-onshore regulatory hostility to "outbound" brokerage; ADR delisting risk; HK/global retail trading cycle exposure.

Verdict — Skip: Real growth + CSRC penalty + regulatory hostility = wrong wrapper for current cycle.

TIGR — UP Fintech

Q1 2026 (reported May 28, 2026): revenue $154.9M (+26.3% YoY). Funded accounts 1.28M; total client assets $58.9B (+28.4% YoY). Net asset inflows $2.9B Q1. Operating income $47.6M (+17.5%); operating margin 34.8%. Net loss $26.9M vs $30.4M income PY — also CSRC regulatory penalty.

Verdict — Skip: Same CSRC overhang as FUTU. Cheaper but same wrong wrapper.

---

Layer 5 — China biotech + healthcare

ONC — BeOne Medicines (formerly BeiGene)

STRUCTURAL DE-RISKING — CRITICAL:

• Renamed BeOne Medicines (formerly BeiGene); effective May 27, 2025
• Ticker changed to ONC (was BGNE) on Nasdaq, January 2025
• Redomiciled from Cayman Islands to Basel, Switzerland via Swiss "continuation" process; effective May 27, 2025; shareholder approved April 28, 2025
• Listings: NASDAQ ONC + HKEX 06160 + SSE 688235 — triple-listing maintained
• The Swiss redomiciliation is the most significant structural de-risking of any Chinese ADR in the cohort — non-China, non-Cayman jurisdiction with mature biopharma regulation

Q1 2026 (reported May 6, 2026): total revenue $1.5 billion (+35% YoY).

• BRUKINSA (zanubrutinib, BTK inhibitor): global revenue $1.1B (+38% YoY); US revenue $761M (+35% YoY) — taking market share from AbbVie's Imbruvica + AZN's Calquence
• TEVIMBRA (tislelizumab, PD-1): global revenue $206M (+20% YoY)
• GAAP diluted EPS $1.96/ADS; non-GAAP $3.24/ADS. Operating income GAAP positive — inflection achieved

Guidance: FY 2026 raised — total revenue $6.3-6.5B; GAAP operating income $750-850M.

Forward P/E ~28-35× FY27e.

Verdict — Core buy: The structural de-risking + commercial inflection + revenue acceleration combination is unique in the cohort. Trade BABA risk for ONC quality. Cleanest "ex-China-ADR-risk China healthcare exposure."

EH — EHang

EHang Holdings — Q1 2026 release scheduled June 9, 2026 (not yet out as of writing). FY 2025 actual: revenue RMB 509.5M (~$72.9M, +11.7% YoY); 221 aircraft delivered (record). EH216-S CAAC Type Certificate + Production Certificate + Standard Certificate of Airworthiness — first eVTOL with all three globally. Routine ticketed commercial operations began March 2026.

Cross-reference defense + space guide for full eVTOL comparison.

Verdict — Hedge: First eVTOL with full CAAC trifecta is unique. Pre-profitability + China geopolitical risk caps sizing.

---

Layer 6 — China ETFs (the diversified wrapper)

ETF	Strategy	Top holdings	AUM
KWEB (KraneShares CSI China Internet)	Pure-play China internet	BABA, TCEHY, PDD, BIDU, JD, NTES, BILI	$6.14B
MCHI (iShares MSCI China)	Broad China large + mid-cap	Top 5: TCEHY, BABA, Meituan, BYD, JD	588 holdings
FXI (iShares China Large-Cap)	50 largest HK-listed Chinese	TCEHY, Meituan, BABA, JD, BIDU, NIO	~60 holdings
CQQQ (Invesco China Technology)	China tech	BIDU, BABA, JD, BILI, KWEB-style	177 holdings
CHIQ (Global X China Consumer)	China consumer	BYD, NIO, YUMC, PDD, JD	58 holdings

Verdict — KWEB or MCHI: Cleanest diversified wrapper; eliminates single-name delisting risk; the recommended exposure for most Indian retail who want China but can't size individual names.

---

What not to chase

• ZK (Zeekr) — DELISTED December 22, 2025 (Geely take-private)
• HSAI (Hesai Group) — named in Select Committee on CCP delisting letter; covered in automotive guide for lidar; ADR risk concentrated
• FUTU + TIGR — CSRC penalties; ongoing regulatory hostility
• VIPS, WB, HTHT — defensive but lower growth than mega-caps
• Speculative micro-caps (BEKE, KE, NIU, etc.) — better China exposure via mega-caps + KWEB

---

Three model portfolios

Pass (recommended for most Indian retail)

0% China ADR allocation. Mega-caps you already own (AAPL, GOOGL, MSFT, NVDA) have indirect China exposure via supply chain + revenue. The structural delisting + VIE + tariff risks aren't compensated by the AI-rerating upside for retail-sized allocations.

Defensive: ETF wrapper + ONC + YUMC + NTES

Bucket	Weight	Names
Diversified ETF	40%	KWEB 40%
Non-VIE healthcare	20%	ONC 20%
Non-VIE consumer	15%	YUMC 15%
Defensive gaming	15%	NTES 15%
Cash	10%	—

Caveat: Total China allocation in your portfolio should still be 1-3% maximum. This is the internal China-bucket allocation.

Balanced: AI thesis + ETF + select mega-caps

Bucket	Weight	Names
China AI mega-cap	35%	BABA 18%, BIDU 12%, JD 5%
Diversified ETF	25%	KWEB 25%
Non-VIE healthcare	15%	ONC 15%
Travel + consumer	15%	TCOM 8%, YUMC 7%
Cash buffer	10%	—

Total China allocation in full portfolio: 2-5% max.

Aggressive: concentrated AI + sum-of-parts plays

Bucket	Weight	Names
China AI mega-cap	50%	BABA 25%, BIDU 15%, JD 10%
Cash-rich value	15%	PDD 15%
Non-VIE healthcare	10%	ONC 10%
Consumer	10%	YUMC 5%, BILI 5%
eVTOL speculative	5%	EH 5%
Cash	10%	—

Total China allocation: 5% max — and acknowledge delisting tail.

---

The risk scenarios

HFCAA forced delisting. PCAOB has complete inspection access since Dec 15, 2022. No Board determinations currently in effect. But: Trump admin Feb 2025 memo flagged VIEs; SEC Chair Atkins publicly stated increased Chinese-issuer enforcement; House Select Committee letters (May, July, Dec 2025) demanded SEC delist named companies (BABA, BIDU, HSAI, ZK). "Two-bad-years-in-a-row away from delisting" overhang — politically driven, outside investor control.

VIE structure invalidation. Every major China ADR is Cayman holding consolidating Chinese operating entity via contractual VIE. Structure never explicitly recognized as legal under Chinese law. Operative + accepted, but holders own "offshore shell with contractual rights" not the actual Chinese entity.

Take-private delisting (ZEEKR precedent). Geely took ZK private at $26.87/ADS Dec 22, 2025. Other state-linked or VIE-vulnerable names could follow.

CAC cybersecurity review (DiDi precedent). July 2021 DiDi forced into delisting after Chinese cybersecurity review post-IPO. Could repeat for any data-sensitive ADR.

US-China tariff escalation. Truce extended through Nov 10, 2026. Effective rate ~39% average. SCOTUS ruled Feb 20, 2026 that IEEPA cannot be used for broad tariffs — Section 301 untouched. Tariff path beyond Nov 2026 is the central unknown.

Currency / capital control. USD/CNY managed band; any depreciation pressure could trigger restrictions.

Consumer demand cycle. China consumption deflation persistent; PPI negative for extended period; instant-commerce subsidy war compressing internet platform margins.

---

How an Indian-resident investor actually executes this

Platform map. US-listed Chinese ADRs (BABA, PDD, JD, BIDU, NTES, etc.) broadly available on Vested / INDmoney / IBKR / Rovia. HKEX dual-primary listings (BABA 9988, others) require IBKR / Rovia primary; some platforms restrict.

Conversion mechanics. If HFCAA forced delisting hits, US-listed ADR holders have theoretical conversion to HK shares for dual-primary names. Conversion fees, timing, broker support vary widely. IBKR best HK access of major Indian retail platforms. Verify your broker before initiating positions.

Tax on gains. Foreign equity LTCG 12.5% under Section 112 with 24-month holding period.

Schedule FA. Calendar-year disclosure for any held during January-December. See Schedule FA guide.

Dividend WHT. US-listed ADRs: US 25% under DTAA. Hong Kong listings: 0% WHT — structural advantage if you can access HKEX directly. Switzerland-domiciled ONC: Swiss 35% reducible to 10% under DTAA via standard W-8BEN — net effective rate on dividends ~10%.

Currency exposure. Long USD on US-listed ADRs; long HKD on HK direct listings. The 24-month LTCG holding period locks in that exposure.

---

The closing

The honest read on June 2026:

• China AI is the cleanest thesis — BABA Qwen3.7-Max #1 Chinese / #5 global, BIDU GPU Cloud +184% YoY, ERNIE 5.1 at 6% of comparable-model pre-training cost.
• HFCAA forced delisting is "two-bad-years away" — politically driven, outside investor control. House Select Committee on CCP has named BABA, BIDU, HSAI, ZK in delisting demand letters.
• ZEEKR delisted Dec 22, 2025 via Geely take-private — one of three delisting pathways (HFCAA, DiDi cybersecurity review, take-private).
• BeOne (ONC) Swiss redomicile May 27, 2025 is the cleanest structural de-risking in the cohort.
• Most major ADRs are HK dual-primary — conversion theoretically possible if HFCAA hits, but mechanics vary by broker.
• 0% China allocation is defensible for Indian retail not specifically convinced on AI thesis or seeking explicit emerging-market diversification.

If you take China exposure: own BIDU + BABA for AI; ONC for ex-China-ADR-risk healthcare; YUMC for non-VIE consumer; NTES for cash-returning gaming. Use KWEB or MCHI ETF for diversified wrapper if you can't size individual names. Cap total China allocation at 3-5% of equity portfolio maximum.

The headlines will continue to oscillate between "China AI is winning" and "Chinese ADRs face delisting."

The Qwen benchmark scores and the PCAOB Board determinations are what actually clear into share prices.