The AI trade has a complicated capital system where money moves around thanks to non-traditional credit. This system works as long as cash flow can pay the bills. Problems arise when cash flow drops, forcing the need for term sheets to cover costs. Oracle’s big investment shows the shift from cash-driven operations to those relying on debt. This change affects how capital cycles work. Moving from relying on equity to engaging creditors indicates a shift in focus. This dependence on private credit could create instability if credit conditions alter, meaning significant impacts on AI capital expenditure (capex). Challenges go beyond just funding; infrastructure issues, especially regarding energy needs, play a big role. New data centers need a lot of electricity, which can’t be quickly solved with money. Special tariffs raise total costs, and delays in energy supply make matters worse. This situation reflects a broader economic cycle where the idea of self-funding productivity is questioned against larger economic factors. The potential AI bubble hinges on the increasing costs of both financial and energy expansion. Financial markets will likely swing between fear and recovery. Key indicators of trouble will not be market feelings but rather tighter funding and setbacks in capex projections, especially as utilities struggle more to meet energy demands. The AI capital spending cycle, where profits just flow back into the system, is starting to show real weaknesses. We are noticing the shift from easy cash flow to debt, and people are seeing the costs of that debt. For example, recent filings from major private credit funds for Q3 2025 show that borrowing costs for new data center projects have risen by at least 75 basis points since the summer. This anxiety is evident in the options market, where risk for the tech sector is now higher than it’s been since the 2024 market panic. Implied volatility on long-dated puts for semiconductor ETFs has risen steadily in the last quarter, indicating traders are hedging against a slowdown. The high growth momentum seen over the past two years is clearly losing strength. Besides the cost of borrowing, we’re reaching a hard limit on electricity. Major utilities are now openly acknowledging the strain. Reports from grid operators indicate that the wait times for new energy-hungry projects extend into 2029. The latest forecast from the Energy Information Administration, released last month, increased its estimate for data center power usage, which current grid infrastructure cannot meet on time. As a result, we are shifting our focus from pure growth strategies to relative value trades that account for these new limitations. We are starting pairs trades by buying call options on manufacturers of transmission and electrical equipment while also buying puts on specific data center REITs in regions facing power shortages. This approach aims to benefit from the physical bottleneck that financial solutions cannot address. The key indicators to watch now are no longer product launches but the details in quarterly reports and the tone of utility earnings calls. We will be listening for any mention of “capex rationalization” from major cloud providers in their upcoming forecasts. This will signify that funding is tightening and the era of growth at any cost is hitting a barrier.

here to set up a live account on VT Markets now