Hyperscale Data Center Financing: Powered Shell and Build-to-Suit Guide 2026

Hyperscale Data Centers: Scale That Requires Institutional Capital

Hyperscale data centers are not typical commercial real estate projects. These facilities are designed to serve major cloud providers including Amazon Web Services (AWS), Microsoft Azure, Google Cloud, and Meta. A single hyperscale campus typically spans 100 megawatts (MW) of power capacity or more, with some campuses reaching 500 MW or beyond. To put this in perspective, a 100 MW data center campus can occupy 500,000 to 1 million square feet and house tens of thousands of servers.

The capital requirements for these projects are staggering. A single hyperscale campus can cost anywhere from $50 million for smaller installations to $500 million or more for large, fully developed campuses with redundant systems, extensive cooling infrastructure, and multiple buildings. These numbers far exceed typical commercial real estate financing and demand specialized institutional capital sources.

Developers and operators pursuing hyperscale projects must understand that traditional commercial mortgage brokers and lenders are not equipped to finance these assets. The financing sources, underwriting criteria, and deal structures are fundamentally different from office, industrial, or retail real estate. This is where specialized expertise becomes critical.

Powered Shell vs Fully Built-Out: What Lenders Prefer

One of the most important concepts in hyperscale data center development is the "powered shell" structure. In this model, the developer finances and constructs the shell building, the power infrastructure (substations, switchgear, distribution), and the cooling systems (chillers, cooling towers, hot aisle containment). The tenant, typically a hyperscaler like AWS or Google Cloud, then brings in its own IT equipment, network infrastructure, and custom fit-out.

This approach offers significant advantages for both developers and lenders. First, it separates the developer's responsibilities from the highly technical and proprietary equipment choices of the hyperscaler tenant. Second, it allows the hyperscaler to maintain control over its competitive advantage: the servers, networking hardware, and software that power its services. Third, it reduces the timeline and complexity of construction because the tenant is not waiting for the developer to install specialized equipment that only the tenant can efficiently deploy.

From a financing perspective, institutional lenders strongly prefer the powered shell structure. Why? Because it creates a cleaner risk profile. The developer is responsible for delivering a functional shell with guaranteed power and cooling capacity. The tenant's credit quality and long-term lease commitment are what drive the deal economics. The lender does not have to worry about whether the developer made optimal choices in server deployment or IT infrastructure, which would be far outside the lender's expertise.

The powered shell approach also makes it easier for lenders to value and stress-test the asset. A fully built-out data center with specific equipment introduces technological obsolescence risk and makes it difficult for a lender (or a future buyer) to reposition the asset if the original tenant leaves. A powered shell with the core infrastructure in place can theoretically serve multiple hyperscalers with different equipment preferences, improving the asset's flexibility and resilience.

Financing Sources for Hyperscale Projects

Hyperscale data center financing comes from highly specialized sources. Traditional commercial banks and mortgage brokers rarely lead these deals because the capital requirements and technical complexity exceed their appetite and expertise.

• Debt funds specializing in infrastructure and data center assets
• Life insurance companies with long-term capital and appetite for 10 to 20 year leases
• Insurance companies seeking stable, inflation-protected cash flows
• Real Estate Investment Trusts (REITs) with dedicated data center debt funds
• Institutional lenders focused on mission-critical infrastructure

These lenders typically deploy larger checks than traditional lenders, often $50 million to $200 million or more per deal. They also have teams dedicated to understanding power procurement, utility interconnection, and the specific credit and operational profiles of hyperscale tenants. They are comfortable with long development timelines because they understand that power interconnection queues can stretch five to eight years in some markets.

Underwriting Criteria

Institutional lenders underwriting hyperscale data center projects apply rigorous, specialized criteria that differ significantly from traditional commercial real estate analysis.

For construction financing, the loan-to-cost (LTC) ratio typically ranges from 55 to 65 percent. This means that for a $100 million construction project, the developer must contribute $35 to $45 million in equity. These conservative LTC ratios reflect the specialized nature of the asset, the tenant concentration risk, and the importance of an anchor tenant letter of intent (LOI) or pre-lease agreement. Most institutional lenders will not fund construction without at least a signed LOI from the tenant, confirming that the campus meets the tenant's specifications and timeline requirements.

For stabilized projects (fully leased and operational), loan-to-value (LTV) ratios are also conservative, typically 55 to 65 percent. This reflects the reality that hyperscale tenants, while credit-worthy (they are often investment-grade companies), represent significant concentration risk. If a single tenant occupies 80 to 100 percent of the campus, the lender's return is entirely dependent on that tenant's creditworthiness and its ability to meet lease obligations over a 10 to 20 year term.

Debt service coverage ratio (DSCR) requirements are typically 1.35x to 1.50x minimum. This means that the net operating income (rent minus operating expenses) must cover debt service by at least 35 to 50 percent. Given that hyperscale leases are typically triple-net (NNN), the tenant pays most operating expenses, which improves cash flow stability and DSCR.

Pre-lease requirements are strict: institutional lenders typically require 50 to 100 percent pre-leased before funding construction. For smaller campuses or single buildings, a 100 percent pre-lease is standard. For larger multi-building campuses, some lenders may accept 50 to 75 percent pre-leased if there is strong demand visibility and the developer has a track record.

Key underwriting elements include:

• Power commitment: Confirmation from the utility that a specific MW capacity is available and the timeline for interconnection
• Interconnection timeline: Detailed utility reports showing when the project can actually receive power
• Tenant credit: Investment-grade credit rating or equivalent strength from the anchor tenant
• Campus redundancy: Whether the campus is designed for N+1 or 2N redundancy (meaning it can operate if one power supply or cooling system fails)
• Fiber access: Confirmation that multiple fiber routes and carriers can serve the campus, avoiding single points of failure

Power Procurement: The Key Constraint in 2026

Power procurement is the single biggest bottleneck in hyperscale data center development. In 2026, utility interconnection queues in key markets remain extensive. Northern Virginia (Loudoun County), a tier-one data center market, has experienced queue times of seven to eight years. Dallas, Phoenix, and Atlanta have similar challenges, though some markets are improving slightly as utilities expand capacity.

This means that for a developer starting a hyperscale project in 2026, the actual power delivery might not occur until 2032 to 2034. A responsible developer must secure a power commitment from the utility early in the process, obtain detailed interconnection studies, and build this timeline into financial projections and lease negotiations.

Lenders scrutinize power procurement heavily. They want to see direct correspondence with the utility, confirmed queue positions, and detailed engineering studies showing the cost and timeline for interconnection. Power is not secondary to the building; in many ways, power is the asset. Land plus power equals value. The building is almost secondary.

Market Outlook: AI and Cloud Drive Demand

Looking ahead to 2026 and beyond, demand for hyperscale data center capacity is expected to remain strong. Artificial intelligence workloads, cloud migration, and the expansion of edge computing all point to continued growth. Major cloud providers continue to announce large capital expenditure commitments, and many are seeking long-term lease arrangements to lock in capacity and pricing.

The most active markets for hyperscale development include Northern Virginia (Loudoun County), Dallas/Fort Worth, Phoenix, Atlanta, Columbus, and Chicago. These markets offer combinations of available land, power infrastructure potential, fiber access, and proximity to population centers and existing cloud provider networks.

Developers and investors with expertise in land acquisition, utility relationships, and cloud provider engagement are well-positioned for the next five years. Those with existing power commitments or positions in markets with shorter interconnection queues have significant competitive advantages.

Contact CLS CRE at 310.708.0690 or loans@clscre.com to discuss financing for your data center project.