Major market maker predicts unprecedented debt issuance as AI infrastructure demands reshape credit markets and corporate balance sheets.
The race to build artificial intelligence infrastructure is about to reshape corporate debt markets in profound ways. According to AI Weekly, Citadel Securities has projected that companies will issue more than $500 billion in new debt between now and 2028, virtually all of it dedicated to purchasing the specialized semiconductors required to power AI systems and data centers.
This forecast carries significant weight because of its sheer scale. Jeff Eason, the head of Citadel Securities' investment-grade desk, estimates that this chip-related borrowing would represent more than 5 percent of the entire Bloomberg US high-grade bond index by 2028. For context, that would make AI infrastructure financing one of the largest single drivers of new credit issuance in the American market.
Why This Matters for Credit Markets
The projection underscores how capital-intensive the artificial intelligence buildout has become. Companies racing to deploy large language models, train neural networks, and operate inference systems need access to expensive processors like GPUs and TPUs. These chips command premium prices, and the global shortage of advanced semiconductors has only intensified competition among technology firms, cloud providers, and enterprises seeking to establish AI capabilities.
The $500 billion estimate reflects a fundamental shift in corporate financing priorities. Instead of issuing debt for traditional expansions or acquisitions, companies are now borrowing hundreds of billions specifically to acquire computing hardware. This reallocation of capital has cascading effects throughout credit markets, potentially influencing interest rates, credit spreads, and the overall composition of investment-grade debt outstanding.
The Implications for Tech Companies and Investors
- Major cloud infrastructure providers will likely drive the majority of new borrowing as they compete to build the most capable AI data centers.
- Semiconductor manufacturers themselves may see increased demand for credit facilities to expand production capacity.
- Credit investors will face new considerations around the timeline and obsolescence risk of AI-specific infrastructure investments.
- The bond market could experience shifts in sector composition and credit quality distributions as AI-focused borrowing accelerates.
The five-year timeline is particularly noteworthy. Eason's projection suggests that the debt issuance will be front-loaded and concentrated, meaning capital markets will experience waves of AI infrastructure financing rather than a gradual increase. This concentration could test the absorptive capacity of debt markets and potentially influence pricing dynamics.
Looking Ahead
The forecast raises important questions about sustainability and return on investment. When $500 billion flows into chip purchases over five years, investors and analysts will scrutinize whether companies can generate sufficient returns to service that debt. The profitability question looms large: can AI applications generate revenue quickly enough to justify the extraordinary capital expenditures?
Citadel Securities' analysis reflects a conviction that artificial intelligence development will continue accelerating and that the hardware requirements will remain insatiable. Whether this prediction proves accurate will depend on the pace of AI adoption, competitive dynamics in the chip market, and broader economic conditions. For now, credit markets should prepare for a historically significant bond issuance cycle centered entirely on the infrastructure of artificial intelligence.
This article was originally published on AI Glimpse.
Top comments (0)