For the first time since 2005, the majority of fund managers surveyed by Bank of America said companies are spending too much on capital expenditure.
Bubbles are notoriously difficult to identify. Predicting when they will pop is even harder. As capital expenditures at tech companies balloon, more investors are considering the possibility that revenue generated from AI investment might not live up to the hype.
Bank of America's latest survey of fund managers found that the majority believe companies are overinvesting and should improve their balance sheets, as seen in the chart below. Those fears have likely been magnified by the scale of investment in AI data centers, costly infrastructure that depreciates over time.
Those fears have likely been magnified by the scale of investment in AI data centers, costly infrastructure that depreciates over time.
Firms making significant AI bets are issuing more bonds to finance their investments. Last month, Blue Owl issued bonds to finance announced a $27 billion joint venture with Meta to build out the company's Louisiana data center.
As the AI buildout expands beyond what Big Tech companies can finance from their own pocketbooks, debt and private equity players have expanded their own investments in the space.
But will this massive collective expenditure -- requiring some $5 trillion, per JP Morgan estimates -- pay off?
If investor confidence continues to wane, that could put a damper on private market AI valuations -- as when the tech hype tumbled in late 2022. Any potential bubble bursting would have far-reaching effects: "I think no company is going to be immune, including us," said Alphabet CEO Sundar Pichai in a BBC interview this week.
For now, however, VCs are still raising, and spending, huge war chests at record valuations without blinking an eye.