Bridgewater Associates founder Ray Dalio recently warned that the AI bubble may be nearing its burst as interest rates rise and tech companies borrow heavily to build AI infrastructure. The situation in the AI sector is indeed becoming increasingly uncertain. As of the end of September, OpenAI's annualized revenue was approximately $50 billion, $20 billion lower than the previously expected $70 billion. With growing market concerns that AI demand may slow and revenue may not be able to support massive infrastructure investments, US tech stocks and the Nasdaq 100 index have come under pressure. This week, selling pressure also spread to semiconductor and tech stocks in Asian markets.
Looking ahead, the high yields on US Treasury bonds could exacerbate the situation, making riskier AI stocks even less attractive to investors. However, this is not a typical bubble, because the massive infrastructure spending is supported by real growth in revenue, orders, and contract backlogs.
For example, Microsoft Azure's annual revenue has exceeded $100 billion, AWS is experiencing its fastest growth in years, and Celestica and AMD are also seeing strong demand for AI infrastructure. In other words, unlike typical bubbles, these companies are not betting on demand that may never materialize. They are struggling to keep up with demand that is already exceeding capacity. Meanwhile, tech giants remain highly profitable and have ample cash reserves to support record investments without seriously jeopardizing their financial stability. However, risks remain: these massive investments have yet to yield returns, and high valuations make the sector vulnerable to underperformance due to lower-than-expected growth. This is a crucial factor to consider when formulating investment strategies. The sector's next major test will be earnings season. This week, in addition to JPMorgan Chase, Goldman Sachs, Wells Fargo, Citigroup, Bank of America, Morgan Stanley, and BlackRock, ASML and TSMC will also release their earnings reports, allowing us to better assess whether the AI boom has the staying power.
