Data Center Development Costs & Market Growth
One union Times
Participating in the development of new data centers eliminates many of the risks associated with purchasing existing facilities, particularly the "lemon" problem, which involves selecting an asset that is difficult to maintain, and the obsolescence risk. The main risk issue is lessened by relying on renewable energy and operating more efficiently. The development sponsor will reap the full benefits of the lengthy first-generation lease because modern data centers growth projections are more likely to attract large credit tenants. At the time of exit, new developments will probably attract a larger pool of institutional buyers. Both Technology Realty Trusts have driven growth through new development rather than pushing rents on existing centers. Sizable development costs (USD 10-14 million/MW of size, according to our evaluation) also create a barrier to entry and require scale to keep fast with the growth of cloud service providers, especially as AI-driven workloads increase.
Many models for data center development exist, each with advantages and disadvantages:
Powered land: The investor acquires land, completes rights, secures the provision of power, and then sells the spade-ready site. The investor avoids working responsibilities and risks around oldness and tenancy—but also misses out on recurring cash flows and long-term upside in evaluation.
Hyperscale build-to-suit: This is the most common development strategy at the moment. Under contract for a long-term lease (frequently 15 years or more), the developer builds an indicated data center to the exact specifications of a major user (likely Big Tech). Advantages include a tenant in place and a long lease term to a credit tenant. Risks center around work and meeting the exacting requirements of Big Tech residents.
Hyperscale or colocation speculative development: Speculative development of a single-tenant hyperscale data center takes advantage of the current strong demand and the potential rent upside from a bidding war, but it runs the risk of not meeting the user's unique technical requirements and/or delivering during a time when demand is slowing. The single-tenant risk and extremely specific technical requirements of hyperscale development are avoided by developing smaller data center automation that serve multiple clients. Little speculative development is currently underway.
Development costs and tenant flexibility
Due to factors like regulatory obstacles, supply chain and uptime disruptions, and labor shortages, rising design and development costs and extended project timelines are becoming increasingly common. These problems necessitate flexibility from businesses and tenants seeking new data center capacity. Tenants are being required to meticulously plan their capacity and backup requirements as the use of colocation, long-term leases, and pre-commitments grows in popularity. This shift necessitates a more strategic approach to capacity planning and risk management from both data center worker and their customers.
Understanding data centers as assets
Continued innovation and adoption of new technologies are powering the increasing digitization of our daily lives: a large part of activities, from communications and banking to marketing and entertainment, have moved to digital sources. The volume of data will rise exponentially from the current levels as this trend continues and AI use grows. Data center buildings that house difficult information technology edge infrastructure crucial for data processing, storage, and high-powered computational services—will have various and necessary roles to play.
Three general types of assets
Many different types of data center assets, but the market can be simplified into three main groups:
Enterprise: Owned and worked by the companies they support and usually housed on the corporate campus; these assets are not available to outside investors.
Colocation: Multiple tenants share a data center. In a retail model, tenants rent racks or cabinets in a data center as a service contract such as power consumption and cooling. Wholesale colocation aspects longer rent than retail, typically five to 20 years, and each tenant pays for power delivery.
Hyperscale: These modern high-capacity data centers were built to meet the massive higher technical and security requirements and pricing demands of the largest technology businesses.
The data center industry is expanding fast, but the factors that are driving that growth are less well understood. Our work, using the proprietary Global Data Center Model, sheds light on the forces underpinning demand for data centers, the unparalleled levels of capital investment underway, and the global dynamics shaping this developing industry.1
Demand for computing in data center's growth forecast is on the rise. GenAI's role in the growing thirst for computing power receives the most attention right now, but the reality is more complex. From 2023 to 2028, we anticipate that the global demand for data center power will grow at approximately 16% on a compound annual basis, which is 33% faster growth than from 2020 to 2023.2 By 2028, the demand will reach approximately 130 GW.
The investing landscape in the data center market
Hyperscalers: Enterprises are pushing to combine AI and machine learning tools into their tech stacks to run core businesses, which will benefit hyperscalers. According to Schneider's writing, "the hyperscalers' ability to create a return on this compounded level of capex has become increasingly top of mind for investors" as we progress from the "infrastructure layer" to the "AI model/source stages" and, ultimately, the "application level." Asset managers: Given a longer time sphere to fully develop data center edge infrastructure, companies that can raise old bridge of private capital will have a possible advantage related to investors with shorter time horizons.
Utilities: One of the most significant potential future constraints on data center supply is the expansion of the power grid. Data center operators: Major data center operators that serve large enterprise customers as well as hyperscale cloud companies and have significant asset footprints will be well-positioned to meet the anticipated global demand surge.