China's Telecom VATS Opening: Who Benefits from History's Pivot Point

Executive Summary

This is not about Vodafone, AT&T, or any telecom carrier. China's April 2024 pilot policy lifts foreign ownership caps on value-added telecom services (VATS) — not basic telecommunications. The distinction is everything.

Basic telecom — mobile networks, broadband fiber, the physical pipes — remains firmly closed to foreign ownership. What is opening: internet data centers (IDC), content delivery networks (CDN), online data processing (EDI), domestic multi-party communications, and information services. These are the services that sit above the network layer.

Since the program launched in October 2024, 166 foreign-funded enterprises have received approvals across the four pilot zones: Beijing, Shanghai, Hainan, and Shenzhen. As of the first batch in February 2025, 13 were specifically telecom businesses.

The analytical error most analysts make: listing which global companies have characteristics that make them "fit" the opening, then concluding they're the beneficiaries. That reverses cause and effect. The right question is: what demand does this policy unlock, and who is positioned to meet it?

1. What the Policy Actually Does

The Negative List and What Changed

For two decades, foreign investment in Chinese telecom was capped at 50% for value-added services. This meant any foreign cloud provider, data center operator, or SaaS company had to find a Chinese joint-venture partner — and operate under that partner's license. The arrangement created persistent tension: the foreign partner wanted global-standard operations; the Chinese partner held the license and, effectively, the keys.

On April 10, 2024, the Ministry of Industry and Information Technology (MIIT) issued a circular removing that cap — for the first time allowing 100% foreign-owned enterprises to apply for VATS operating licenses. The pilot formally launched on October 23, 2024.

What's Inside the Pilot (and What's Not)

The Four Pilot Zones

The policy applies in four designated areas: Beijing, Shanghai, Hainan, and Shenzhen. Critically, a company approved in any of these zones can serve customers nationwide — the geographic restriction is on the registration, not the customer base.

2. Why Demand-Driven Analysis Is Correct

The Error: Feature-Matching Instead of Demand-Tracing

The naive analysis goes: "Here are the attributes of companies that would benefit from China opening its telecom market. Companies matching those attributes win."

This is backwards. The policy doesn't create benefit because certain companies have certain features. It creates benefit because it unlocks demand — demand that was previously suppressed by regulatory friction. The companies that win are simply the ones positioned to meet that demand when it's released.

Two Sources of Demand, One Policy Trigger

Demand Source 1: Inbound

Multinational corporations operating in China have long wanted to use the same cloud infrastructure, collaboration tools, and data services they use everywhere else. But under the old 50% cap, foreign cloud providers had to operate through Chinese joint ventures — arrangements that created persistent concerns about data sovereignty, IP protection, and operational control. Many multinationals either accepted degraded service (using local alternatives with different APIs, security models, and compliance postures) or avoided putting sensitive workloads in China entirely.

With 100% foreign ownership now permitted, these providers can offer their services directly, under their own operating licenses. The compliance path is clearer. The data control question — who holds the keys? — has an unambiguous answer. For multinational customers, this removes the single largest objection to deploying their full technology stack in China.

Demand Source 2: Outbound

Chinese companies expanding overseas — the Shein, Temu, TikTok, BYD generation — need global cloud infrastructure, CDN capacity, and enterprise tools to serve users outside China. They've been using these services, but often through workarounds or gray-zone arrangements, because the foreign providers lacked a legitimate license to serve them from within China.

Now, with foreign service providers holding proper Chinese operating licenses, these outbound companies can contract with them from within China, under Chinese law, with clear compliance. The arrangement is no longer a regulatory workaround; it's a sanctioned business relationship. For risk-averse enterprises, this is the difference between "we can probably do this" and "this is fully approved."

3. The Market: Size and Dynamics

China's IDC and Cloud Market

China's internet data center market was valued at approximately $38.85 billion in 2023, projected to reach $77.37 billion by 2029 at a 12.1% CAGR. The broader data center market — including servers, networking, and infrastructure — was estimated at $30.5 billion in 2024, growing to $121.6 billion by 2035.

Cloud Market Concentration

The Chinese cloud market is dominated by domestic players: Alibaba Cloud (Alibaba), Huawei Cloud, and Tencent Cloud control the majority of market share. AWS and Azure operate through Chinese partners — AWS through Sinnet and NWCD, Azure through 21Vianet. Google Cloud has no China region at all.

4. Who Benefits: The Three Categories

Two demand streams — inbound multinationals and outbound Chinese firms — converge on providers that can serve both. The beneficiaries fall into three categories, each addressing a different layer of the value stack.

Category 1: Global Cloud Service Providers

These are the primary beneficiaries. Cloud is the foundation layer. Inbound multinationals need global-consistent cloud architecture; outbound Chinese firms need the same to reach international users. With the ability to hold their own VATS licenses, the big three global clouds can now offer compliant infrastructure to both customer sets from within China.

Why they benefit:

• AWS — Already has physical infrastructure in China (Beijing, Ningxia) but operated through partners. A WFOE license lets them bring operations under direct control, addressing the data sovereignty concerns that prevented many multinationals from using AWS in China. Also serves as the natural cloud platform for Chinese outbound companies that already use AWS globally.
• Microsoft Azure — Similar position. Azure China operated by 21Vianet. A WFOE license lets Microsoft offer direct Azure services — critically, with the same security and compliance tooling (Entra ID, Purview, Sentinel) that enterprise customers use globally. This is the single most requested capability from multinational CTOs with China operations.
• Google Cloud — No existing China region. This is both a disadvantage (starting from zero) and an advantage (no legacy JV to unwind). Google's strengths in AI/ML infrastructure (Vertex AI, TPUs) and data analytics (BigQuery) are precisely the capabilities that both inbound and outbound enterprises most want — and that domestic clouds are weakest at delivering.

Category 2: Global Data Center Operators

These are secondary beneficiaries, following the cloud providers. When a global cloud provider establishes a WFOE in China and needs physical infrastructure, they don't build every data center from scratch. They lease from colocation providers. And they strongly prefer global-standard operators — the same companies they work with in Singapore, Frankfurt, and Northern Virginia.

Why they benefit:

• Equinix — The world's largest colocation provider, already present in Shanghai through a joint venture. A WFOE structure lets Equinix expand its China footprint to serve cloud providers and enterprise customers directly, offering the same Equinix Fabric interconnect that its global customers expect.
• Digital Realty — Similar dynamics. Global enterprise customers that use Digital Realty facilities worldwide want the same operational standards, security certifications, and interconnection capabilities in China. A WFOE license makes this possible.

Category 3: Enterprise Collaboration and Communication Tools

These are tertiary beneficiaries, enabled by the cloud and IDC layers above. Collaboration tools don't need their own data centers — they run on cloud infrastructure. But they do need operating licenses to provide domestic multi-party communication services in China. Under the old rules, they couldn't get those licenses without a Chinese JV partner. Now they can.

Why they benefit:

• Zoom, Microsoft Teams, Webex — Video conferencing and team collaboration tools that are standard in multinational enterprises. With the policy opening, these platforms can deploy China-resident infrastructure and hold their own operating licenses, providing the data residency and compliance guarantees that enterprise customers require.
• Slack, Notion, Monday.com — Productivity and workflow tools face the same dynamic. Their enterprise customers with China operations want to use the same tools everywhere. A compliant China deployment — with data stored locally, under the provider's own license — removes the regulatory objection.

5. Who Does NOT Benefit (And Why That Matters)

6. What We Don't Know Yet

Any analysis of Chinese regulatory opening must account for uncertainty. Several open questions will determine how large the opportunity actually becomes:

1. Licensing speed and predictability. The MIIT approval process for VATS licenses under the pilot remains opaque. 166 approvals in ~20 months is fast by Chinese regulatory standards, but the criteria are not public, and rejections aren't disclosed. The pipeline matters more than the first batch.
2. Data localization requirements. China's data security regime (CSL, DSL, PIPL) still requires certain categories of data to remain within China. A foreign-owned IDC in Shanghai gives you physical data residency, but cross-border data transfer rules — security assessments, standard contracts, certification — still apply when moving data out.
3. Pilot expansion timeline. The pilot is currently limited to four zones. Whether it expands to other cities, and on what timeline, will determine how much of the addressable market is actually reachable. China has a pattern of "pilot, evaluate, expand" — but the evaluation period is unpredictable.
4. Geopolitical risk. The policy exists in the context of US-China technology decoupling. Export controls on advanced semiconductors (October 2022, updated 2023–2025) restrict what hardware foreign cloud providers can deploy in China. The policy opens the legal door; export controls may limit what can walk through it.