Companies are spending record amounts on AI, but that doesn’t mean that these boom times will last forever.
Ashwath Damodaran, the NYU Stern finance professor widely regarded as the world’s foremost authority on valuation, has a warning for anyone riding the AI wave: history says a correction is coming, and when it does, the pain could be significantly worse than what the world experienced during the dot-com crash. The difference, he argues, lies in how the current boom is being financed — and how much physical infrastructure is being built on the back of that capital.
Damodaran starts with an important distinction. “I don’t know whether there’d be a bust, but history suggests there would be a bust,” he says. But whatever form a correction takes, he argues it would hit harder than the dot-com bust of the early 2000s. His reasoning begins with capital expenditure. The dot-com boom, he points out, was not driven by heavy infrastructure spending. “There was no huge capital expenditure in that cycle. In fact, there was very little traditional CapEx or even R&D driving it. People started apps, they started basically going online.” That era was, by nature, asset-light. The current AI buildout is anything but.
“This has been the biggest infrastructure run-up I’ve ever seen of a business,” Damodaran says, going so far as to compare it to the automobile industry’s expansion a hundred years ago. “The amount of money that’s being put into AI CapEx is immense, which means that when the correction comes, the pain will be more intense.” Goldman Sachs estimates that total hyperscaler capex from 2025 through 2027 will reach $1.15 trillion — more than double everything spent in the three years prior. That is the scale of what Damodaran is referring to.
The second, and arguably more consequential, difference is how this boom is being funded. The dot-com era ran almost entirely on equity. “The dot-com boom and bust was almost entirely equity funded,” Damodaran notes. “When the bust came, there were shareholders who lost 60, 70, 80 or even 90% of the money. You felt sorry for them, but the loss was restricted to the shareholders.” Painful as that was for investors, the damage stayed contained. What is different now is that the AI CapEx boom is not being financed through equity alone. “The problem with the AI CapEx boom is not only is it immense, but a big chunk of it is funded with debt, the debt coming from private capital rather than banks.”
That last detail matters. Debt from private capital markets — rather than traditional bank lending — tends to be less transparent, less regulated, and harder to unwind in a crisis. “There’s a very real chance that if there’s a correction and companies start having problems, that problem is going to show up as distress and default. And that pain doesn’t stay restricted. It spills over into the rest of society.”
Damodaran is careful not to invoke 2008 casually, but he does invoke it. “I’m not saying it’s going to be 2008, but 2008 is an example of what happens when lenders overreach, when they lend money at too low a rate. And the correction comes, the pain spills over.” His core concern is about a broader societal cost: “the potential societal cost of having to deal with debt coming due that you’re unable to pay.”
It’s a concern that others in finance are beginning to share. Zoho founder Sridhar Vembu has called AI “clearly an investment bubble”, noting that the profits from AI remain largely on paper while the CapEx is very real. Damodaran himself has previously said the AI market would need trillions in revenues to justify current valuations — and has acknowledged that “collectively, the pricing just seems too rich.” He has also exited his position in Nvidia entirely. None of this means the technology itself is worthless, or that the AI buildout will necessarily end in a crash. But Damodaran’s framework offers a useful lens: the bigger the infrastructure bet, and the more of it funded through debt, the more damage a correction can cause — well beyond the investors who made the original wager.