Oracle logo /Courtesy of Yonhap News

Recently, as the "artificial intelligence (AI) bubble" theory has resurfaced, sales have increased for derivative financial products with an insurance-like nature that pay compensation if corporations default. With the race to invest in AI overheating, investors appear to be hedging in advance to prepare for a potential burst of the bubble.

The Financial Times (FT), citing data from the U.S. Depository Trust & Clearing Corporation (DTCC), reported on the 15th that credit default swap (CDS) transaction volume linked to U.S. tech corporations rose about 90% from September to mid-month.

A CDS is a derivative in which compensation is paid if corporations fall into debt default, and it is a key indicator showing how investors perceive the credit risk of those corporations.

FT said CDS transaction volume increased as investor anxiety grew over some tech corporations that issued large-scale bonds to raise funds for massive AI infrastructure investments. AI investments such as data centers, high-performance semiconductors, and cloud facilities require astronomical capital, but it can take years to translate into actual revenue, which is weighing on the market.

This trend was pronounced at cloud service providers Oracle and CoreWeave, which raised billions of dollars in liabilities for AI facility investments. Facebook parent Meta also saw related CDS transactions surge after issuing $30 billion in bonds in Oct. this year to fund its AI business.

CDSs linked to major tech corporations saw little demand early this year when AI optimism was the mainstream. But as third-quarter results for tech corporations such as Oracle and Broadcom recently fell short of expectations, more investors sought CDSs, FT reported. AI-related sales were weak relative to the scale of investment.

In the second half of this year, U.S. big tech corporations such as Meta, Amazon, Google parent Alphabet, and Oracle secured $88 billion (about 129.7 trillion won) for AI projects. U.S. investment bank JPMorgan projected that AI-related funding raised by investment-grade corporations will approach about $1.5 trillion (about 2,210 trillion won) by 2030.

A senior official at a major U.S. bond investment firm said in an interview with FT, "As more people seek to use CDS portfolios on big tech corporations or Oracle and Meta, single-name CDSs have increased noticeably," adding, "In the current environment, the most common way to hedge and protect assets is to set up a tech stock CDS basket."

The AI bubble theory has surfaced whenever the stock market has swung, but it is drawing renewed attention as big tech's AI investments mount. The slowdown in AI tech stocks' growth is also fueling the bubble theory. The AI industry's structure is such that share prices rise and funds flow in based on forecasts of explosive growth, but if major corporations' results are weaker than expected, that belief can be shaken.

In results for Sept.–Nov. released last week, Oracle's cloud infrastructure sales and cloud sales both missed market expectations, sending the stock plunging and triggering a surge in corporate bond selling. FT reported that Oracle's CDS price soared to a record high since 2009. This is interpreted to mean there has been a sharp increase in demand to hedge Oracle's risk.

Shares of U.S. AI Semiconductor corporation Broadcom plunged about 11.4% in a single day on the 11th–12th on assessments that its future order backlog was disappointing. Nvidia also did not escape the headwinds from the AI bubble theory, falling 5.7% over the past week (the 8th–12th).

※ This article has been translated by AI. Share your feedback here.