The rise of AI has created a situation where expected future earnings are being used to finance today’s infrastructure. Initially, big companies were reinvesting their profits, but now Oracle has taken a different path by borrowing heavily, with a debt-to-equity ratio of 500%. The industry needs $500 billion each year for capital investments, leading to a growing dependence on debt to support AI projects.

Venture Credit Boom

Venture capital has evolved into venture credit, with banks and private funds stepping in to finance this change. It is estimated that around $3 trillion will be spent on global data centers by 2028, with half of that funding coming from outside sources. This results in a $1.5 trillion shortfall, leading to private credit and “AI infrastructure bonds” being issued for institutions looking for returns. AI-focused corporations now make up 14% of the U.S. investment-grade bond index, surpassing the banking sector. However, these companies don’t have the same financial backing and regulations as banks, putting the credit market at risk. AI technology requires a significant amount of power, with capital needs reaching half a trillion dollars yearly. If new technologies disrupt current systems, the inflated equity could burst, impacting associated debts. AI is now expensive, similar to industrial commodities like oil, but it doesn’t generate its revenue. The credit market supports hope through investments based on projected revenues instead of actual earnings. We’ve entered an era where substantial borrowing fuels optimism for AI. Without real productivity improvements, the sector may inflate asset values instead of driving meaningful innovation. The sustainability of the debt used to unlock AI’s potential is still in question. The market appears to be misjudging the core issue of the AI boom, which is likely found not in stock valuations but in credit markets. A vast amount of debt is being used to build infrastructure based on revenue forecasts that could take years to realize. This creates a significant gap between perceived risk and the actual leverage in the system.

Future Credit Events

We are witnessing signs of complacency as of early October 2025. Credit default swap spreads for key data center operators and AI-related hardware companies have hardly budged this year, despite a historic $450 billion in bond issuance. This indicates that the market sees these firms as low-risk giants instead of ventures relying on the uncertain success of future technologies. For traders in derivatives, this suggests buying protection against an upcoming credit event. A straightforward strategy is to purchase longer-dated credit default swaps (CDS) for the most heavily leveraged players in the AI supply chain. Some firms are borrowing heavily to support large, long-term energy and infrastructure projects, with debt maturing in just three to five years. A more comprehensive strategy involves buying put options on investment-grade and high-yield corporate bond ETFs. AI-related debt now constitutes over 14% of the U.S. investment-grade index, a higher concentration than the entire banking sector. If AI profitability is reassessed, this could quickly widen credit spreads, directly affecting these funds’ values. This credit vulnerability will eventually impact the equity market, potentially in a dramatic way rather than just leading to corrections in valuations. Long-dated put options on major hyperscalers or the Nasdaq 100 index can be a good way to protect against this risk. Triggers for downturns could include regulatory restrictions on energy use or breakthroughs that render billions in current GPU infrastructure outdated overnight. This situation feels reminiscent of the telecom infrastructure bubble between 1999 and 2001, where companies heavily invested in fiber optic cables based on speculative demand. When that demand didn’t arrive as expected, the equity collapsed, but the real damage occurred in the credit markets that backed the dream. We now see pension funds and insurers being sold “AI infrastructure bonds” based on similar hopes. Create your live VT Markets account and start trading now.

here to set up a live account on VT Markets now