TLDR: Morgan Stanley forecasts a significant $1.5 trillion funding gap in global AI data center infrastructure by 2028, which traditional hyperscaler cash flows cannot fully cover. The firm predicts that credit markets, encompassing both public and private sectors, will become increasingly vital in bridging this deficit, with substantial contributions expected from private credit, corporate debt, and securitized markets.
The burgeoning capabilities of generative artificial intelligence are driving an unprecedented capital investment cycle, primarily centered on the expansion of data centers. According to a recent report from Morgan Stanley, this surge in investment spending on data center construction and power generation is projected to contribute up to 40 basis points to U.S. real GDP growth between 2025 and 2026.
Morgan Stanley anticipates global data center spending to reach approximately $2.9 trillion by 2028. This colossal sum is broken down into roughly $1.6 trillion for hardware, including chips and servers, and $1.3 trillion dedicated to building the essential data center infrastructure, which encompasses real estate, construction costs, and ongoing maintenance. Annually, investment requirements related to data centers are expected to exceed $900 billion by 2028.
Despite the substantial cash flows generated by hyperscale cloud service providers, these alone will not suffice to meet the escalating investment demands. Morgan Stanley’s equity analysts project that while approximately $1.4 trillion of hyperscaler capital expenditure may be self-funded through internal cash flows, a considerable $1.5 trillion financing gap will remain.
Vishwanath Tirupattur, Morgan Stanley’s Global Head of Quantitative Strategy, emphasized the pivotal role of credit markets in addressing this shortfall. “The key takeaway from the report is that credit markets—secured, unsecured, and securitized in both public and private markets—will play a growing role in financing data centers,” Tirupattur stated. He added, “We think that credit markets, broadly defined to encompass both public and private markets of different flavors, will gain traction as more efficient providers of capital to bridge this gap.”
The report details how various credit avenues are expected to contribute to closing this funding deficit. Private credit is anticipated to provide a significant portion, potentially up to $800 billion. This will be complemented by approximately $200 billion from corporate debt, $150 billion from ABS/CMBS (Asset-Backed Securities/Commercial Mortgage-Backed Securities) markets, and an additional $350 billion from sovereigns, private equity, and banks. Tirupattur highlighted that private capital, particularly private credit, is optimally positioned to meet a large part of the remaining financing gap, given its alignment with the expansion of Assets Under Management (AUM) in a higher interest rate environment and the complex, customized financing needs of AI infrastructure build-out.
Also Read:
- India’s Data Centers Rapidly Scale Up to Meet Surging AI Computing Demand
- Artificial Intelligence Set to Drive Singapore’s Sustained 3% Annual GDP Growth
The scale of AI investment is unprecedented, with the total capital expenditure of all S&P 500 index companies in 2024 projected to be around $950 billion, underscoring the immense financial requirements of the AI-driven data center boom. This macroeconomic impact is expected to make data center financing a key long-term focus for credit investors.
