AI Capital Expenditures Continue to Rise, Tech Giants' Credit Spreads Widen, Financing Costs for Companies like Alphabet (GOOGL.US) Under Pressure

According to Zhitong Finance APP, as investment in artificial intelligence infrastructure continues to heat up, investor concerns over massive capital expenditures by tech giants are spreading from the stock market to the bond market. Recently, the bond credit spreads of major cloud service providers such as Alphabet (GOOGL.US), Amazon (AMZN.US), and Meta (META.US) have continued to widen, reflecting that fixed-income investors are demanding higher risk compensation to address the financing pressure and uncertainty of returns brought by AI investment.

This week, after Alphabet announced a further increase in its full-year capital expenditure forecast, the company’s bond yields rose, sparking market worries that Microsoft (MSFT.US), Amazon, and Meta may also continue to ramp up AI investment in the future.

Market participants have pointed out that one key driver behind the sustained growth in capital expenditures is the continuous rise in energy costs. Currently, major cloud providers are building AI data centers across the US at an unprecedented pace, and electricity has become one of the most significant operational costs for such projects.

GE Vernova CEO Scott Strazik stated that, against the backdrop of ongoing geopolitical tensions, inflationary pressures are expected to remain elevated. This week, the international oil price again surpassed $100 per barrel, further intensifying concerns over rising data center operating costs.

Meanwhile, the continued rise in US Treasury yields has also made fixed-income investors more cautious.

Mizuho Securities noted in its latest report that the steadily increasing capital expenditure budgets are gradually approaching the upper limit of investor tolerance. Tech giants, once regarded as “fortresses of capital,” are now facing the pressures of rapidly escalating AI-related costs.

Mizuho expects that by next year, the overall capital expenditure scale of major cloud service providers will surpass their total free cash flow for the first time, which means these companies will become increasingly reliant on external financing to support AI infrastructure construction.

A credit fund portfolio manager stated that currently, bond investors and equity investors are engaged in intense debate over AI investments. The market’s focus has gradually shifted to whether these massive capital expenditures can generate sufficient returns.

As concerns continue to rise, Oracle's (ORCL.US) five-year credit default swap prices have once again reached multi-year highs.

Barclays analyst Andrew Keches stated in a report released this week that Oracle CDS has once again become an important hedging tool for market participants against AI financing risks. It not only reflects the company's fundamentals but also represents concerns over AI capital expenditures, OpenAI project execution, and the overall outlook for data center investment.

As both the construction and leasing of data centers continue to expand, investors are now paying closer attention to whether Oracle will continue to increase its debt financing scale in the coming years.

Earlier this month, S&P Global downgraded Oracle’s credit rating to BBB-, just one notch above junk status.

However, Oracle’s management remains optimistic about the company’s AI strategy. In recent years, the company’s cooperation with OpenAI has deepened, and it has worked with Meta, Nvidia (NVDA.US), and others to jointly promote the construction of AI cloud infrastructure.

Analysts believe that the continued rise in tech companies’ bond financing costs could also impact the financing of large AI projects in the future.

Sources revealed that Meta is currently preparing to finance its Texas data center project, with total investment of around $12 billion, and the pricing for this financing is expected to be finalized early next week. According to media reports, the financing rate for this project is expected to be higher than for Meta’s similar past projects, indicating that investors are not only demanding greater risk compensation but are also beginning to pay more attention to whether AI infrastructure investments can deliver matching returns.