Since the arrival of OpenAI’s ChatGPT in late 2022, the list of possible indicators that the AI boom was about to explode is long and growing, leaving Wall Street wondering each time if the end is finally near. For example, every time AI chip giant Nvidia saw its market cap surpass the $2 trillion, $3 trillion, $4 trillion, or $5 trillion milestones, the bears sounded the alarm. Excess borrowing by AI hyperscalers and their astronomical capex plans raised more red flags. The DeepSeek shock in early 2025 sparked another panic when the Chinese AI model raised questions about the need for all that capital spending. Also last year, OpenAI CEO Sam Altman’s admission that an AI bubble exists and Michael Burry’s bearish bets represented more red flags. This year, the Iran war, the collapse of the South Korean stock market and “circular” financing deals between AI companies added to fears. But profits continued to rise and the stock eventually recovered to set one record after another. Of course, traders’ persistence in continuing to bet on AI could also be a sign of a bubble. Many on Wall Street also recognize that the good times will come to an end at some point. The only question is how long AI bets can be placed without leaving money on the table. Now there’s a new reason to wonder if it’s time to cash out. The current angst over AI centers on whether the trillions of dollars spent on AI will pay off sufficiently, dragging down debt-laden companies like CoreWeave. Computer company neoclound saw its shares fall and credit default swaps rise recently. For Dan Greenhaus, strategist at Solus Alternative Asset Management, CoreWeave is not a canary in the coal mine indicating that the downfall of AI is imminent. “How many canaries have we killed in the coal mine over the last three years as everyone has latched onto every new sign of weakness, so to speak, as evidence that the bull market was about to end only to watch the broader market, both tech and otherwise, reach new highs?” he told CNBC on Friday. “So I don’t know if I would read more into what’s happening now than I would have read into the other headlines.” To be sure, borrowing costs have risen as the Iran war and oil prices pushed bond yields to the highest levels in more than 20 years. The Federal Reserve is also raising rates again to curb persistent inflation. This has put greater scrutiny on the riskiest areas of the AI boom, which has become increasingly reliant on issuing debt through the bond market. But Greenhaus noted that CoreWeave’s unsecured debt has been especially attacked, calling it “the riskiest of the riskiest AI development plays.” Meanwhile, earnings from AI chip suppliers such as Taiwan Semiconductor indicate that the broader demand environment remains strong. At the same time, he also does not expect that companies in the AI ecosystem will be able to continue funding themselves indefinitely as rates rise. “I don’t know if this is the breaking point or if another 25 or 50 basis points more is the breaking point,” Greenhaus explained. “But as rates go up, the more speculative tails, so to speak, of construction are going to be challenged, and right now that’s CoreWeave.” For his part, CoreWeave CEO Mike Intrator said markets can “misinterpret and act irrationally” and told employees in a memo that the company’s backlog of deals is more than enough to sustain its debt obligations. Nvidia CEO Jensen Huang also backed CoreWeave, which counts the chipmaker as a shareholder, saying it does not need financial backing. “CoreWeave’s strength is improving every day,” he told Bloomberg. “I’m optimistic. Their cost of capital is actually going down.” Others on Wall Street also see strong demand for AI and think the recent sell-off is another buying opportunity. Still, even surprising revenue estimates may not be enough to support sky-high spending. On Wednesday, Gartner warned that AI data center owners must generate cumulative revenue of $8.5 trillion through 2030 to cover capital spending, financing costs, and a 12% return. But total enterprise and consumer AI revenues heading into data centers will only be $3.6 trillion at most, exposing more than $4 trillion in asset value to risk of write-downs through 2032. On Friday, Bank of America analysts said in a note that the AI investment boom is following a classic capital cycle, typically ending in disappointment “even though the underlying technology often lived up to expectations.” Typically, initial profit margins tend to be high, leading to overinvestment and excess capacity while attracting more competition. In the case of AI, competition has already increased, especially from Chinese open weight models. On top of that, enterprise spending on AI has started to fall as higher token volumes fail to offset lower token prices. “This monetization disappointment in the face of intense competition is precisely the mode of failure seen in past capital cycles,” BofA added.
