The $200 Billion Borrowing Binge
The race to dominate artificial intelligence has shifted from a battle of algorithms to a battle of balance sheets. Tech giants and data center developers have unleashed a historic borrowing spree, issuing over $200 billion in AI-related debt this year alone. As hyperscalers like Microsoft, Meta, and Google rush to secure computing power, the sheer volume of capital required, estimated at up to $10 trillion for the full infrastructure rollout, is beginning to distort credit markets.
Reports from [fortune.com](https://fortune.com/2025/12/12/ai-data-center-boom-sparks-fears-of-glut-amid-lending-frenzy/) indicate a market in a state of near-frenzy, with some developers seeking loans covering 150% of construction costs, banking on future valuations rather than current cash flow. While equity markets celebrate the AI boom, the debt markets are flashing warning signs of saturation and concentration risk.
The “Hidden” Leverage
Not all of this debt is visible on public balance sheets. A growing trend involves financial engineering designed to mask the true extent of corporate leverage. As noted by [nytimes.com](https://www.nytimes.com/2025/12/15/technology/ai-risks-debt.html), major tech firms are increasingly utilizing off-balance-sheet structures and “synthetic leases” to finance data centers. By renting computing power from smaller intermediaries who take on the primary debt, these giants effectively push the financial risk down the food chain to smaller lenders and private upstarts.
This “toothpaste tube” effect, where squeezing risk in one area simply forces it out elsewhere, complicates the picture for investors. Transparency is diminishing just as capital demands are hitting their peak, creating a precarious environment where the stability of the underlying infrastructure assets becomes difficult to assess.
Concentration Limits and the Credit Squeeze
The public bond market has physical limits. Institutional portfolios often have strict caps on single-sector or single-issuer exposure. With the technology sector consuming a massive share of new issuance, [forbes.com](https://www.forbes.com/sites/brandonkochkodin/2025/11/19/trillion-dollar-ai-borrowing-binge-could-spark-the-next-credit-crunch/) reports that diversification rules are forcing lenders to turn away new paper, regardless of the borrower’s quality. This saturation threatens to crowd out other borrowers, potentially widening spreads for non-AI companies as liquidity is vacuumed up by the tech sector.
Bond Capital: Hard Assets Over Hype
While the headlines focus on the speculative nature of AI software, the underlying reality is a massive physical build-out. Data centers are ultimately real estate and utilities projects. They require steel, concrete, and power grids.
For borrowers, the message is clear: public markets may be entering a period of indigestion. At Bond Capital, we look past the speculative valuation of the technology itself and focus on the tangible assets. The physical build-out of AI requires massive, flexible capital that traditional banks, constrained by concentration limits, may struggle to provide. While equity grabs the headlines, disciplined private debt finances the concrete. We remain focused on lending against verifiable asset value, ensuring that even if the digital bubble deflates, the physical infrastructure retains its worth.
