Morgan Stanley believes the rumor of China's RMB 2 trillion data center investment may be a misreading; industry view remains In-Line

Overview

This is an event-driven commentary on China's cloud and data center industry. The report focuses on market concerns about rumors of potential large-scale Chinese government investment in data centers, discussing their authenticity, policy implications, and impact on the competitive landscape for data centers, public cloud, and state-owned telecom operators. Morgan Stanley believes the rumor has not been officially confirmed and may have misread policy language about investment in data center interconnection networks and related infrastructure upgrades as direct new data center construction.

Core views

The core views are: first, news reports about RMB 2 trillion of data center investment over the next five years need to be interpreted cautiously, because existing official policy appears more aimed at promoting upgrades to data center interconnection, networks, computing power, and security facilities. Second, the main bottleneck in China's AI development is not data center capacity supply, but chip production and design. Third, public cloud share continues to tilt toward hyperscale cloud vendors such as Alibaba and ByteDance, due to better chip supply chains, stronger model capabilities, and faster product innovation, and the report does not believe state-owned telecom operators will reverse this trend. Fourth, if the report does indeed point to large-scale data center construction led by state-owned telecom operators, it could cause clear oversupply and hurt data center operators.

Analysis framework

The report uses an event-driven and policy-text cross-checking approach: it first compares the news report with the RMB 2 trillion investment wording in the official January 2025 policy document, and then assesses the impact on the industry from the perspectives of industrial bottlenecks, supply-demand policy, cloud vendor competitiveness, and potential supply shocks.

Methodology notes

• Event-driven researchCross-checking news reports with official policy

  Cross-validate the investment scale in media reports against the semantics of official policy documents.

  The report points out that the RMB 2 trillion in Bloomberg's report may come from the January 2025 policy document, but that document places greater emphasis on the construction and upgrading of networks, computing power, and security facilities between data centers, rather than direct data center construction.

• Industry supply-demand analysisData center supply-demand balance assessment

  Assess whether new government-led investment would lead to data center oversupply.

  The report believes the current policy orientation is toward balancing data center supply and demand; if state-owned telecom operators build data centers on a large scale, it could run counter to that direction and cause significant oversupply.

• Competitive landscape analysisComparison of hyperscale cloud vendors and state-owned telecom operators

  Explain changes in public cloud market share from the perspectives of supply chain, model capability, and innovation speed.

  The report believes Alibaba and ByteDance have stronger advantages in chip supply chain, model capability, and rapid innovation, and therefore continue to gain public cloud market share from state-owned telecom operators.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• BABA

  A hyperscale cloud vendor; the report says it continues to gain share in public cloud relative to state-owned telecom operators.

  Strengths

  Better access to supply chains, including imported and local chips as well as in-house T-head; has Qwen model capabilities; relatively fast innovation speed.

  Weaknesses

  The report does not provide specific financial or valuation weaknesses.

  Comparison

  Relative to state-owned telecom operators, Alibaba has stronger advantages in chip supply chain, model capability, and innovation pace.

  Risks

  If the government pushes state-owned telecom operators to build and operate data centers on a large scale, public cloud competition may intensify.

• ByteDance

  An unlisted hyperscale cloud and AI participant; the report says it is also gaining public cloud market share.

  Strengths

  Has model capabilities such as Seedance and strong rapid innovation capability.

  Weaknesses

  It is unlisted, and the report does not disclose investable securities or valuation information.

  Comparison

  Relative to state-owned telecom operators, ByteDance is stronger in models and innovation.

  Risks

  Policy-driven expansion of state-owned infrastructure could change the competitive environment.

• SOE telcos

  Potential builders and operators of government data center investment.

  Strengths

  May receive policy and capital support and possess basic communications network resources.

  Weaknesses

  The report believes they are being surpassed by hyperscale cloud vendors in public cloud market share, and their model capability, supply chain, and innovation speed may lag internet cloud vendors.

  Comparison

  Relative to BABA and ByteDance, state-owned telecom operators are at a disadvantage in public cloud competitiveness.

  Risks

  If they undertake large-scale construction, it could cause industry oversupply and depress returns.

• data center players

  Industry assets directly affected by potential government-led expansion.

  Strengths

  AI demand growth remains a long-term source of demand.

  Weaknesses

  If the scale of new investment is too large, oversupply would weaken pricing and utilization.

  Comparison

  Compared with public cloud platforms, independent data center companies are more sensitive to supply expansion and utilization changes.

  Risks

  If the RMB 2 trillion data center construction is true, it would be a major negative factor.

Key data

• Rumored investment scaleus$295bn / rmb2 trillionBloomberg reported that the Chinese government may invest this amount over the next five years for nationwide data centers.
• Corresponding wording in policy documentrmb2 trillion over 5 years / rmb400bn per yearThe report believes this figure is more likely derived from the January 2025 policy document, whose original text points to data infrastructure driving the construction and upgrading of networks, computing power, and security facilities.
• Report date2026-06-12The time on the report cover page is June 12, 2026 04:04 AM GMT.
• Industry viewIndustry View In-LineThe report shows the industry view for Greater China Telecoms and Asia Pacific as In-Line.
• Rating time frame12-18 monthsMorgan Stanley states in its disclosures that industry views and stock ratings generally correspond to the next 12 to 18 months.

Impact & implications

In terms of investment implications, the report reduces the market's immediate concern about large-scale direct government construction of new data centers, because the related report has not been officially confirmed and may involve a semantic misreading. For the cloud computing and data center sectors, the near-term key question is whether policy is really shifting toward expansion led by state-owned telecom operators; if not, the share advantage of hyperscale cloud vendors may continue. If the rumor is confirmed, then the risk of data center oversupply would rise, and both independent data center companies and the public cloud competitive landscape could come under pressure.

Risks

• If Bloomberg's report is officially confirmed, large-scale data center investment led by state-owned telecom operators could trigger significant oversupply.
• The market may misread investment in data infrastructure interconnection as direct data center construction, causing short-term valuation volatility.
• China's AI infrastructure is still constrained by bottlenecks in chip production and design, and data center expansion may not solve the core bottleneck.
• Public cloud competition may change if policy resources tilt toward state-owned telecom operators.
• The report includes regulatory disclosures and potential conflict-of-interest warnings, and investors should not rely solely on a single research conclusion.

What to watch

• Whether the National Development and Reform Commission or other official channels confirm the RMB 2 trillion investment plan.
• Whether subsequent policy documents clearly distinguish between investment in data center interconnection networks and investment in direct data center construction.
• Whether state-owned telecom operators receive new data center construction mandates, capital expenditure guidance, or cloud business support policies.
• Changes in supply-demand balance, rack utilization, and pricing in the data center industry.
• Changes in public cloud market share of hyperscale cloud vendors such as Alibaba and ByteDance.
• Changes in China's AI chip supply, domestic chip design, and availability of imported chips.