Morgan Stanley Sees $2.9T Data Center Spending, $1.5T Gap
Morgan Stanley forecasts global data center spending of about $2.9 trillion from 2025 to 2028, driven by AI infrastructure demand. The projected outlay includes $1.6 trillion for chips and servers and $1.3 trillion for construction and other physical infrastructure.
Amazon, Microsoft, Alphabet, Meta and Oracle are expected to fund about $1.4 trillion from cash flows, leaving an estimated $1.5 trillion financing gap. Private credit could provide roughly $800 billion, while hyperscaler lease commitments exceed $1.2 trillion in undiscounted future payments. Nvidia has reportedly discussed backstop guarantees of up to $250 billion for data center debt, but no agreement is confirmed.
Financing is not the only constraint: the US data center power shortfall is forecast to reach 32–33 GW by 2028. For traders, the scale of data center spending highlights both sustained AI-related demand and growing reliance on debt. Credit conditions, hyperscaler commitments and power availability could affect AI-sector valuations and broader risk appetite.
Neutral
This is primarily a technology-sector financing story, not a direct cryptocurrency catalyst. In the short term, the forecast may support sentiment around AI-linked equities and infrastructure, but the large financing gap, reliance on private credit and power constraints also highlight execution and credit risks. Crypto traders may watch these developments as part of broader risk appetite, since digital assets have at times moved alongside technology stocks and liquidity-sensitive assets; the article, however, provides no direct evidence of an immediate effect on crypto prices or flows.
Past periods of heavy investment in emerging technologies have often brought both strong growth expectations and concern about overbuilding. If AI spending continues and credit remains available, the buildout could sustain confidence in growth-oriented markets over the longer term. If borrowing costs rise, projects are delayed, or AI demand and productivity gains disappoint, investors could reassess valuations and reduce exposure to risk assets, including crypto. Relevant indicators include credit spreads, hyperscaler capex and lease commitments, project timelines, power availability, and broader liquidity. On the information presented, the likely crypto-market impact is indirect and uncertain, so a neutral classification is appropriate.