Western AIDC Development Drives Profit Growth and ROE Re-rating for Chinese IPPs

Overview

This report focuses on investment opportunities for independent power producers (IPPs) arising from artificial intelligence data center (AIDC) development in Western China. UBS believes the market has not fully priced in the high-return project opportunities enabled by large-scale Western AIDC construction. It forecasts that captive AIDC projects in Western China will generate an annual profit pool of approximately RMB 16 billion between 2026 and 2030, driving 13–25% earnings growth for covered IPPs by 2030. The report upgrades China Resources Power, China Power International Development (CPID), and Xinyi Energy to Buy, maintains Longyuan Power at Buy, and significantly raises target prices.

Core views

The core thesis centers on the "Compute-power Coordination Policy," which UBS believes will convert Western China’s resource advantages into tangible returns for IPPs. Demand side: China’s AI demand has reached an inflection point, with daily token consumption hitting 140 trillion in March 2026—up 1,400x from 100 billion in early 2024. Strong AI demand, sustained capex by domestic cloud providers, and GPU supply constraints imply that newly built AIDC capacity in Western computing hubs (e.g., Inner Mongolia, Gansu, Ningxia, Guizhou) will quickly achieve high utilization. Policy side: The April 2026 release of the "AI-Energy Mutual Empowerment Action Plan" made the Compute-power Coordination Policy operationally actionable for the first time. The policy mandates that over 80% of power for new data centers in national computing hubs must come from renewable sources by end-2025, with direct-supply arrangements emerging as a key implementation pathway. The Datang Zhongwei project, commissioned in May 2026 as the first large-scale grid-connected project, marks the beginning of IPPs bidding for captive renewable projects via direct-supply agreements. Project economics: Captive AIDC projects in Western China offer significantly superior economics versus conventional projects. Assuming a 50% solar+storage / 50% wind+storage mix, project IRR is estimated at ~20%, with unit profit of RMB 0.15/W, compared to only 5–11% IRR and RMB 0.07–0.12/W for conventional new projects. This advantage stems from: (1) 24/7 AIDC load under direct-supply contracts driving higher utilization hours (solar: 1,927 hrs in West vs. national avg. 1,088 hrs; wind: 2,716 hrs vs. 1,979 hrs); and (2) flat terrain, low land lease costs, and economies of scale reducing unit construction costs by 25–30% versus Eastern China. Market share assumptions: Based on historical execution of large projects, balance sheet strength, and regional presence in the West, UBS assumes China Resources Power captures 20% market share by 2030, Longyuan Power and CPID each 10%, and Xinyi Energy 1%. Accordingly, Western AIDC is expected to boost 2030 earnings by 17% (CRP), 22% (CPID), 25% (Longyuan), and 13% (Xinyi).

Analysis framework

UBS employs a DCF (Discounted Cash Flow) model for company valuations, with key assumptions including: risk-free rate of 4.3%, cost of equity of 8.7–10.6%, after-tax cost of debt of 2.2–3.0%, WACC of 6.7–8.0%, and terminal growth rate of 1–3%. Valuation adjustment logic: The target price increases are primarily driven by higher terminal values rather than near-term earnings revisions. Terminal ROIC assumptions were raised to reflect the structural uplift from high-IRR Western projects: CRP from 7.0% to 7.7%, CPID from 7.1% to 7.2%, Longyuan from 7.1% to 7.4%, and Xinyi from 7.0% to 7.7%. Correspondingly, terminal NOPAT multiples were also increased. Sensitivity analysis: Key variables were stress-tested. On installed capacity, if Western AIDC accounts for 55–80% of incremental demand, the 2030 profit pool ranges from RMB 12.4B to RMB 18.0B (base case: RMB 15.8B). On power price discounts, assuming direct-supply discounts of 5–25% vs. wholesale prices, solar+storage IRR ranges from 15–26%, and wind+storage from 15–28%. On utilization hours, ±10–15% variation yields project IRRs of 14–28%.

Methodology notes

• Valuation MethodDCF Discounted Cash Flow

  DCF Valuation

  The report uses a DCF model to estimate intrinsic value by forecasting future free cash flows, discounting them at WACC, and adding terminal value. Notably, terminal ROIC assumptions were raised to reflect the structural improvement in long-term returns from high-IRR Western projects.

• Industry/Sector Analysis FrameworkSupply-demand framework

  Supply-Demand Framework and Utilization Hour Analysis

  Project returns in power generation hinge on utilization hours (generation / capacity). The report quantifies the economic uplift from AI data centers by comparing Western captive project utilization (solar: 1,927 hrs vs. national avg. 1,088 hrs; wind: 2,716 hrs vs. 1,979 hrs).

• Company Fundamentals & Financial FrameworkROIC–WACC spread

  ROIC vs. WACC Spread Analysis

  The report highlights the spread opportunity as project IRR (~20%) significantly exceeds current ROE (5–11%). As high-return captive projects grow in portfolio weight, blended ROIC will trend toward 20%, supporting P/BV re-rating.

Asset mapping & comparison

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

• China Resources Power (0836.HK)

  Top beneficiary, assuming 20% market share. Its strong track record in large projects, robust balance sheet (ample cash), and attractive dividend yield (4%) make it best positioned to execute large-scale Western captive projects.

  Strengths

  Excellent execution track record, strong cash position, attractive dividends, relatively high ROE base (11%)

  Weaknesses

  Conventional thermal power faces pressure from electricity price caps and slowing demand

  Comparison

  Market share assumption is double that of CPID and Longyuan, and 20x that of Xinyi Energy

  Risks

  Higher-than-expected coal prices, rising debt costs, higher construction costs, lower-than-expected plant utilization

• China Power International Development (2380.HK, CPID)

  Benefits from 10% market share assumption. Its low current ROE (6%) implies the highest ROE uplift elasticity from incremental projects, with projected 22% earnings growth—the highest among the four companies.

  Strengths

  Highest ROE uplift elasticity, mature renewable development capabilities, strong Western regional presence

  Weaknesses

  High leverage (net debt/equity 194%), near-term cash flow pressure, potential need for equity financing

  Comparison

  Same market share assumption as Longyuan (10%), but lower ROE base drives higher earnings elasticity

  Risks

  High leverage, cash flow strain during capital-intensive phase, potential equity financing needs

• Longyuan Power (0916.HK)

  Maintained at Buy, assuming 10% market share. As China’s largest wind power operator, it holds a leading position in resource-rich Western regions like Inner Mongolia and Gansu.

  Strengths

  Largest wind developer in China, deep Western regional presence, largest gap between project IRR and current ROE (4x) implies greatest re-rating potential

  Weaknesses

  Concerns over curtailment, delays in renewable subsidy recovery

  Comparison

  Same market share assumption as CPID (10%), but lower ROE base (5%) and pure renewable profile enhance benefit from green power demand

  Risks

  Lower-than-expected wind utilization, reduced policy support, lower-than-expected tariffs, higher curtailment rates

• Xinyi Energy (3868.HK)

  Upgraded to Buy, assuming 1% market share. Despite small share, its limited asset base means even modest contributions significantly improve portfolio returns.

  Strengths

  Marginal improvement highly impactful due to small scale, terminal ROIC raised by same magnitude as CRP (to 7.7%)

  Weaknesses

  Limited Western presence, low appetite for new projects, subsidy recovery delays

  Comparison

  Market share assumption significantly lower than peers (1% vs. 10–20%), reflecting strategic focus away from Western expansion

  Risks

  Slower-than-expected PV capacity expansion, larger-than-expected tariff declines, insufficient subsidy funding

Key data

• 2030 Western AIDC Profit PoolRMB 15.8 billionBased on 29.5 GW capacity and 60 TWh incremental demand, assuming 50/50 solar-wind mix
• China Resources Power Target Price IncreaseHK$16.80 → HK$33.60+100%, based on 20% market share assumption and terminal ROIC raised to 7.7%
• CPID Target Price IncreaseHK$2.60 → HK$5.10+96%, based on 10% market share assumption
• Longyuan Power Target Price IncreaseHK$7.80 → HK$9.30+19%, Buy rating maintained
• Xinyi Energy Target Price IncreaseHK$1.25 → HK$1.57+26%, upgraded to Buy
• Western Captive Project IRR~20%Significantly above current ROE (CRP 11%, CPID 6%, Longyuan 5%, Xinyi 7%)
• Utilization Hour AdvantageSolar: 1,927 hrs vs. national avg. 1,088 hrsWind: 2,716 hrs vs. 1,979 hrs, enabled by 24/7 direct-supply contracts
• Unit ProfitRMB 0.15/WWestern captive projects vs. RMB 0.07–0.12/W for conventional projects
• AI Token Consumption Growth1,400xDaily average of 140 trillion in Mar 2026 vs. 100 billion in early 2024

Impact & implications

The report argues that Western AIDC development represents a new structural growth driver for Chinese IPPs, currently underpriced by the market. High-IRR captive projects will not only directly contribute earnings (starting H2 2026) but more importantly push overall corporate ROE toward 20%, supporting P/BV re-rating. China Resources Power is positioned as the top beneficiary, assuming 20% market share, thanks to its proven track record in large projects, strong balance sheet (ample cash), and attractive dividend yield (4%), making it best suited for large-scale Western captive projects. CPID, despite having the lowest current ROE (6%), offers the highest ROE elasticity from incremental projects, with a projected 22% earnings uplift—the highest among the four. Longyuan Power, as a pure-play renewable operator, exhibits the largest gap between project IRR and current ROE (4x), implying significant re-rating potential. Xinyi Energy, with limited Western exposure (1% share assumption), still achieves meaningful portfolio return improvement due to its small asset base. At the industry level, while the 60 TWh incremental demand represents less than 0.5% of China’s total annual generation—insufficient to trigger a nationwide super-cycle—it presents a rare high-return growth opportunity for individual IPPs. The report expects sector P/E to revert from the current 9.7x toward its historical average of 12.7x as high-return projects materialize.

Risks

• Higher-than-expected coal prices squeezing thermal power margins (for CRP and CPID)
• Rising debt costs impacting project financing
• Higher-than-expected construction costs
• Lower-than-expected plant utilization
• Larger-than-expected declines in renewable tariffs
• Weaker-than-expected AI demand growth reducing AIDC utilization
• Policy implementation delays hindering direct-supply contract roll-out

What to watch

• Initial earnings contribution from Western AIDC projects in H2 2026
• Provincial-level implementation rules for compute-power coordination policies
• Continued growth trajectory of AI token consumption
• Progress on captive project awards to companies
• Outcomes of coal-electricity long-term contract price negotiations (linked to coal price trends)
• Renewable subsidy recovery progress