Sentiment shifted in the stock market this week as investors began to question the value of corporate software and IT services in an age where artificial intelligence may be able to do the job in-house. But the market hasn’t significantly dropped despite the selloff, with analysts saying that investors are instead rotating through to other sectors.
This might feel familiar to the more seasoned investor, as it is a flash of déjà vu to the late 1990s when the rewards and risks posed by dotcom innovations began trickling through.
The AI boom, most analysts agree, is not the same as the dotcom bubble and its subsequent pop. This week’s shift away from IT and SaaS (software as a service) assets was in relation to an update from Anthropic, which launched plug-ins for its Claude Cowork agent last week that could streamline work in data analysis, legal, marketing, and sales.
The fallout has been relatively contained: Neither the S&P 500 or the Nasdaq was down more than 2% yesterday, and stocks across Europe and Asia are relatively flat this morning. But the cycling out of certain impacted sectors, even on a relatively small level, and into other industries, does beg comparison with shifts the market observed in the last technological revolution.
As Deutsche Bank’s Henry Allen highlighted to clients this morning, the software component of the S&P is down nearly 30% from its peak in October, and as such, “you’d be forgiven for thinking markets would have seen a huge correction by now.”
“However, what we’ve actually seen is a significant rotation … Other sectors have taken up the baton from tech, such as energy, materials, and consumer staples, meaning that the overall S&P 500 still only closed –2.6% beneath its record high from last month,” Allen said.
“Interestingly, that pattern echoes what we saw in 2000 as the dotcom bubble started to burst,” he added. “Equities started to fall from March 2000 as tech stocks saw significant declines. However, consumer staples, utilities, and health care rallied significantly over the months ahead, and in September the S&P 500 actually came within a percentage point of its record high from six months earlier.
“So it shows that a market can absorb a prolonged rotation without obvious index-level stress for some time. But the longer and deeper the selloff in a dominant sector becomes, the harder it is for the broader index to withstand the drag, and the continued losses for tech in 2000 ultimately meant the S&P 500 ended that year over –10% lower.”
Where the similarities end
While the Deutsche analyst did highlight that the selloff isn’t easing yet—Amazon slumped on earnings last night with investors digesting $200 billion in AI expenditures—that doesn’t mean a similarity with behavior observed in the dotcom era means the end result will be the same.
For starters, as Eric Sheridan, senior equity research analyst at Goldman Sachs, pointed out in an October report, the notion of an AI bubble coming down the pipe is far more prevalent now than discussions around the dotcom or housing bubbles were when they were in their midst. He said: “It’s true that some characteristics of the current period rhyme with past bubbles … But in 1999, firms that had no revenue were the ones with the most exuberant valuations.
“Today, most of the Magnificent Seven—which trade at an aggregate P/E of 31x versus 23x for the market … generate outsized levels of free cash flow and engage in stock buybacks and pay dividends, which very few firms did in 1999.”
When global GDP is hinged on AI capital expenditures (and bullishness on its future payoff), it is only right that investors would be asking themselves if or when the bet will pay off—and whom it will pay off for. This is the argument from Jamie Dimon, CEO of JPMorgan Chase, who told the Fortune Most Powerful Women Summit in October: “You can’t look at AI as a bubble, though some of these things may be in the bubble. In total, it’ll probably pay off.”
