The Thesis
Digital Realty is a data center landlord that rents out specialized, high-power buildings to the world's largest cloud providers and technology companies. The company generated $6.11 billion in revenue last year, representing 10% growth as it continues to expand its global footprint. The massive wave of investment in artificial intelligence marks the structural shift that is turning these facilities from commodity real estate into the essential infrastructure of the modern economy.
What makes this investment work boils down to four specific things.
In our view, Digital Realty is a high-quality compounder that is currently priced for perfection, given that the market value has run well ahead of historical norms. The case for owning it from here depends on whether you believe AI demand will keep rental prices moving higher for the next decade. For long-term investors, this is the cleanest way to own the physical foundation of the internet.
Numbers at a Glance
What does it do?
Digital Realty is a mature business that earns money by leasing specialized buildings and power to companies that need to store and run massive amounts of computer hardware. When a cloud provider or a large bank needs a home for its servers, it rents space in one of Digital Realty’s 309 facilities. The company provides the cooling systems, high-speed internet connections, and uninterruptible power supply that these machines require to stay online 24 hours a day. Customers pay a monthly rent based on the amount of power they use and the square footage they occupy, typically signing long-term contracts that last several years.
Where does revenue come from?
The vast majority of revenue comes from long-term rental contracts and fees for connecting customers to one another. Rental revenue includes both the physical space and the power used by servers. Interconnection revenue comes from the specialized cables that allow two companies inside the same building to talk to each other at lightning speeds. Tenant reimbursements cover the shared costs of running the facilities, such as taxes and insurance.
Who are its customers?
Digital Realty serves 309 data centers globally, supporting everything from massive cloud providers to individual corporate enterprises. As of March 2026, the company manages 3.0 gigawatts of total IT capacity and has an additional 6.3 gigawatts of buildable capacity waiting in its development pipeline. The portfolio is currently 90.1% occupied, with the top 20 customers providing a significant portion of the company's total annual rent. These customers span across several industries, including financial services, manufacturing, and social networking.
What gives it staying power?
Staying power comes from the high switching costs and the massive scale required to compete in global data center markets. Once a customer installs thousands of servers and links them into Digital Realty's ecosystem, moving them to a different building is expensive and risky. The 234,000 cross-connects between customers create a network effect that competitors cannot easily replicate.
Where is it headed?
The company is shifting its entire focus toward capturing the tidal wave of demand for artificial intelligence infrastructure. Management is aggressively building out 6.3 gigawatts of new capacity to house the specialized, power-intensive chips required for AI training. If successful, this shift moves Digital Realty from being a provider of general-purpose office space for servers to the primary infrastructure partner for the AI revolution.
Revenue growth remains steady as the company successfully raises prices on existing customers and adds new capacity. Total operating revenue reached $1.64 billion in the first quarter of 2026, up 16% from the prior year. This acceleration reflects strong demand for high-density space that can support artificial intelligence workloads.
Free cash flow is often negative because the company chooses to spend billions of dollars every year building new data centers. While the business generates steady cash from its existing tenants, its $2.41 billion in 2025 free cash flow was a rare positive outlier. Investors should view this high spending as a deliberate reinvestment in growth rather than a sign of a struggling business model.
The balance sheet is becoming more resilient as management reduces debt relative to the company's earnings power. Net debt-to-adjusted EBITDA has fallen to 4.7x, down from 5.1x a year ago, showing a commitment to safer financial footing. This leverage is common for real estate businesses, but the current downward trend provides more room for future construction.
Digital Realty is a financially solid infrastructure business that is currently prioritizing massive capital reinvestment to capture the AI market opportunity.
The company is successfully pushing through higher rental rates as the global supply of power-connected land becomes more scarce. Core FFO per share reached $2.04 in the most recent quarter, showing that the company can grow its bottom line even while spending heavily on new buildings. This pricing power is the direct result of having a limited supply of high-voltage data center space in key cities.
Interest rates remain the single biggest external risk because they dictate the cost of funding Digital Realty's multi-billion dollar construction projects. If interest expense rises significantly, it can eat into the profits generated by new data centers before they are even finished. Management is currently managing this by reducing debt levels, but a prolonged period of high rates will slow their growth engine.
The data center industry is worth roughly $250 billion today and is growing at 15% annually, putting it on track to exceed $450 billion by 2029. It is a highly attractive industry because power and land with fiber connectivity are structurally scarce, giving landlords significant pricing power. Digital Realty stands as one of the two dominant global leaders in this market, positioned to capture the massive shift toward AI-specific infrastructure.
Competition for the world's largest tenants is intense, but the high cost of construction and power permits creates a significant barrier to entry. The industry is currently consolidating as established players with cheap access to capital buy up smaller local operators.
Equinix(EQIX) is the primary threat because its massive network of 234,000 connections makes it the "airport" of the internet where everyone must meet. Private equity-backed firms like CyrusOne and Vantage compete aggressively on price for the largest "hyperscale" deals with companies like Google or Meta. The most dangerous threat is the rising cost of power and local utility constraints, which limits how fast new competition can enter a market.
Digital Realty is holding its ground and gaining share in the high-density AI segment by leveraging its massive existing power pipeline.
The primary protection for this business is the high cost of switching and the "efficient scale" of its global footprint. Once a customer is physically connected to hundreds of other companies inside a Digital Realty facility, moving those servers would cause unacceptable downtime and cost. The company’s 234,000 cross-connects act as a physical network that gets stronger with every new tenant.
The company's ROIC of 1.9% is low, reflecting the massive amount of capital required to build these facilities before they become fully profitable. However, the 90.1% occupancy and falling debt-to-EBITDA ratio suggest that the underlying assets are high-quality and durable. The combination of high occupancy and rising rental rates proves that the company has real pricing power in a tight market.
The moat is strengthening as the complexity and power requirements of AI workloads make it harder for smaller competitors to build viable facilities.
Core FFO grew to $2.04 per share, beating internal targets during Q1 2026.
Reduced net debt-to-EBITDA from 5.1x to 4.7x while funding 6.3GW development pipeline.
CEO Andrew Power has a significant multi-million dollar equity stake and performance-based compensation.
Capital Allocation Track Record
Andrew Power has successfully navigated the transition from traditional data storage to high-intensity AI workloads while simultaneously fixing the company's balance sheet. The management team’s decision to lower debt levels while doubling down on 6.3 gigawatts of new capacity shows a rare combination of caution and ambition. This disciplined approach makes them one of the more trustworthy teams in the infrastructure space.
© 2026 ClearThesis.ai · Report generated on May 27, 2026
This is an AI-generated analysis for informational purposes only and does not constitute financial advice. Data and analysis may not reflect recent developments if viewed significantly after the generation date. Always conduct your own due diligence before making any investment decisions.