GDS (GDS) Q4 2025: AI-Driven Bookings Triple to 300MW, Unlocking 10%+ Yield on New Projects
GDS’s fourth quarter capped a transformative year as AI-fueled demand drove new bookings to three times prior levels and set the stage for an aggressive expansion into China’s emerging hyperscale data center hubs. Management’s disciplined capital allocation, robust asset monetization, and a strategic pivot toward AI-centric projects have positioned GDS to scale rapidly while maintaining double-digit investment yields. Investors now face a landscape where AI workload concentration, falling unit pricing, and competitive barriers will define the next phase of growth and risk.
Summary
- AI Demand Reshapes Growth Trajectory: GDS’s pipeline and bookings are now overwhelmingly AI-driven, fundamentally altering its revenue mix and project focus.
- Capital and Landbank Scaled for Expansion: Asset monetization and a RMB 2.8B cash reserve empower GDS to pursue 3GW of new development in national hub markets.
- Yield and Cash Flow Stability: Despite lower unit pricing, disciplined development and monetization keep portfolio yields above 10% and support positive pre-financing cash flow.
Performance Analysis
GDS delivered 11% year-over-year growth in both revenue and adjusted EBITDA, surpassing the top end of guidance and achieving positive pre-financing cash flow for the first time. The company’s asset monetization strategy, including an ABS transaction and CREIT IPO, enabled deconsolidation of project companies and a leaner balance sheet. Pro forma, revenue and EBITDA growth would have reached 13.2% and 14.2%, respectively, if not for these asset sales.
Gross new bookings soared to over 96,000 square meters (300MW), three times the level of the previous three years, with 60–70% of new business tied to AI workloads. Move-ins hit a record 86,000 square meters for the year, while net debt to EBITDA improved from 6.8x to 5.8x—falling further to 4.8x when factoring in recent capital raises and time deposits. Unit pricing (MSR) fell 2.4% year-over-year, reflecting a shift to edge and new market projects, but this was offset by lower development costs, keeping gross portfolio yield steady at 11%.
- AI Bookings Surge: Over 500MW of gross new bookings targeted for 2026, with AI workloads dominating demand.
- Cash Reserves Bolstered: Proceeds from asset sales and convertible preferred shares have lifted cash to $2.8B, supporting aggressive expansion.
- Operational Efficiency Gains: Accounts receivable days fell from 109 to 82, boosting operating cash flow and freeing capital for reinvestment.
GDS’s ability to maintain double-digit yields and positive cash flow, even as pricing drifts lower, signals a robust business model that is adapting to hyperscale AI demand while managing capital intensity and risk.
Executive Commentary
"AI in China has really taken off. Increasing availability of domestic high-performance chips is a key enabler. All our major customers are investing in hyperscale computing infrastructure to support AI adoption. As a result, we are seeing robust recovery in data center demand across both new markets and established markets."
William Huang, Founder, Chairman, and CEO
"Our operating cash flow for the full year was around 3.4 billion RMB. The significant improvement year on year was helped by a reduction in AR days from 109 in 4Q24 to 82 days in 4Q25 as a result of our tight control of collections. After taking into account asset monetization proceeds, we achieved positive cash flow pre-financing of 1 billion RMB."
Dan Newman, Chief Financial Officer
Strategic Positioning
1. AI-Centric Bookings and Pipeline
AI workloads now account for the majority of new business, with management targeting 500MW of gross bookings for 2026 and reporting 200MW already secured plus 500MW of MOUs in early 2026. This shift is driven by hyperscale cloud and large language model (LLM) customers, reflecting a structural change in demand patterns and project delivery timelines.
2. National Hub Expansion and Landbank Accumulation
GDS is building a three gigawatt pipeline focused on China’s official national data center hubs—Horinger (Inner Mongolia), Zhongwei (Ningxia Province), and Shaoguan (Guangdong Province). The company’s landbank has expanded to 3.7GW, providing the platform for multi-year growth and supporting both AI and traditional cloud workloads.
3. Asset Monetization and Capital Efficiency
Through asset sales (ABS, CREIT) and equity raises, GDS has de-risked its balance sheet and secured RMB 2.8B in cash, allowing for self-funded expansion and reduced leverage. The company’s “develop, ramp, monetize” model ensures capital is recycled efficiently, with new investments generating 10–11% cash yields and 20%+ return on equity.
4. Competitive Barriers and Customer Concentration
Entry into new markets is protected by high regulatory and financial hurdles, with government selection favoring operators with proven track records, customer commitments, and capital strength. Customer concentration is intensifying, with the top three AI players in China driving the bulk of new demand and locking in long-term (7–10 year) contracts.
5. Pricing Dynamics and Yield Management
MSR (monthly service revenue per square meter) continues to decline, driven by location mix and market pricing, but lower development costs and scale allow GDS to maintain portfolio yields at 10–11%. Management expects further MSR pressure (3–4% annual declines) but sees volume growth and asset recycling sustaining overall returns.
Key Considerations
GDS’s 2025 performance signals a decisive pivot to AI infrastructure, enabled by disciplined capital management and a focus on national data center hubs. Investors should weigh the implications of concentrated AI demand, evolving pricing, and operational leverage as the company enters its next growth phase.
Key Considerations:
- AI-Driven Growth Profile: Bookings and backlog are now heavily weighted toward hyperscale AI customers, increasing both growth potential and customer concentration risk.
- CapEx and Landbank Readiness: A RMB 9B CapEx plan and 3.7GW landbank set the stage for multi-year expansion, but require flawless execution and continued access to capital recycling.
- Yield Preservation Amid Pricing Pressure: Falling MSR is offset by lower build costs and asset monetization, keeping investment returns attractive for now.
- Competitive Moats in New Hubs: Regulatory and financial barriers provide GDS with a defensible position in China’s emerging data center clusters.
- Operational Discipline: Improved receivables and cash flow management underpin the company’s ability to self-fund growth and manage leverage.
Risks
Customer concentration is rising, with top AI players accounting for most new demand, exposing GDS to volume and pricing risk if these customers slow investment. MSR declines could erode margins if not offset by cost reductions or higher utilization. Regulatory changes, chip supply volatility, and intensifying competition in new markets could disrupt project delivery or returns. The company’s expansion plan also hinges on continued success in asset monetization and timely project execution.
Forward Outlook
For Q1 2026, GDS guided to:
- Continued ramp of AI-driven bookings and move-ins, with 200MW of new orders secured and 500MW of MOUs pending conversion.
- Maintaining portfolio yields at 10–11% despite further MSR declines.
For full-year 2026, management guided:
- Total revenue of RMB 12.4B–12.9B (8.5–12.8% YoY growth).
- Adjusted EBITDA of RMB 5.75B–6B (6.4–11% YoY growth).
Management emphasized that CapEx of RMB 9B is sufficient for current sales targets, and future monetization events are not yet included in guidance. Move-ins are expected to double in 2027 if bookings targets are met this year, supporting a multi-year growth runway.
Takeaways
GDS’s 2025 results confirm a business model transition toward AI-driven hyperscale infrastructure, underpinned by strong capital discipline and a scalable platform in China’s national data center hubs.
- AI Workload Dominance: The company’s future is now tied to the trajectory of China’s AI adoption, with bookings and project delivery increasingly concentrated among a few hyperscale customers.
- Balanced Growth and Risk: Asset monetization, yield management, and operational discipline provide a buffer against pricing pressure and capital intensity, but customer and market risks are rising.
- Watch for Execution in New Hubs: Successful ramp-up and monetization in new markets will be key to sustaining double-digit returns and defending market share as competition intensifies.
Conclusion
GDS enters 2026 with a record backlog, robust cash reserves, and a clear AI-centric growth strategy, but faces a landscape where execution, customer concentration, and pricing dynamics will test its ability to deliver outsized returns. Investors should monitor the pace of AI adoption, asset monetization outcomes, and the company’s ability to maintain yields amid falling unit pricing.
Industry Read-Through
GDS’s performance signals a broader inflection in China’s data center industry, as AI workloads rapidly overtake traditional cloud demand and national hub strategies reshape geographic and capital allocation priorities. The shift to longer-term, hyperscale contracts and the emergence of regulatory and financial barriers in new markets will likely favor well-capitalized incumbents, raising the bar for new entrants. Falling unit pricing, offset by scale and asset recycling, is a trend to watch across the sector, with implications for margin structure and competitive positioning among global and domestic peers. As China’s AI infrastructure buildout accelerates, operators able to balance capital discipline with aggressive capacity deployment stand to capture disproportionate share.