AI stocks could fall sharply even if the technology continues to grow, especially if profits fail to meet investors’ expectations.
Slower spending on chips and data centers, higher borrowing costs, and cheaper competing technology could trigger a correction.
A deeper downturn could develop if disappointing returns lead to spending cuts, tighter lending, and further losses throughout the AI industry.
Ken Griffin, head of Citadel, is the latest Wall Street titan to suggest AI stocks are overpriced.
"Corporate America has been very uneven in its ability to extract value from its AI models," he said in an interview with Bloomberg. He went on to suggest that the Street “is almost certainly” in an AI bubble.
But despite repeated cautionary words from people in the know, AI stocks continue to go higher. So, what would actually cause AI stocks to fall?
The answer may have less to do with whether AI transforms the economy than with how much companies earn from it — and how quickly those earnings arrive.
A correction, conventionally defined as a decline of at least 10% from a recent peak, could begin when investors reduce their profit forecasts or become unwilling to pay such high prices for anticipated growth. Individual companies, semiconductor stocks, and the broader market could experience different degrees of decline.
The International Monetary Fund’s October financial stability report identifies weaker or less durable AI investment and profitability as potential triggers for falling AI asset prices. It also warns that financial vulnerabilities could amplify market turmoil.
Strong growth might not be enough
One potential trigger is a disappointing earnings report from a major AI company. Revenue wouldn’t necessarily have to decline. If investors expect extraordinary growth, a merely strong result could prompt a selloff.
Rising expenses could have a similar effect. Companies might attract more customers while spending so much on computing, electricity, and development that profits fall short of expectations.
Another trigger would be a decision by major cloud providers to slow their infrastructure expansion. Postponed data centers or reduced equipment orders would affect the suppliers counting on that business.
Goldman Sachs has highlighted how AI investment supports supplier earnings, while warning that slower spending growth and increasing depreciation expenses could weaken that support.
When spending fails to pay off
A more fundamental problem would emerge if customers found that AI products didn’t generate enough additional revenue or cost savings to justify their expense.
Businesses could reduce subscriptions, delay new projects, or demand lower prices. That would put pressure on software providers and eventually the companies supplying their computing infrastructure.
Technological progress could also unsettle investors. More efficient models or cheaper competing chips might make AI more affordable while reducing demand for particular equipment or squeezing suppliers’ profit margins.
Similarly, excessive construction could leave the industry with more computing capacity than customers need, pushing prices lower.
Borrowing costs could compound the damage
Higher interest rates could pressure AI stocks in two ways: by making infrastructure more expensive to finance and by reducing the value investors assign to profits expected years into the future.
Businesses dependent on borrowing would be particularly exposed. Construction delays or difficulty obtaining electricity could compound their problems, leaving them paying for equipment and financing before facilities generate revenue.
Recent reporting describes lenders demanding greater compensation to finance some AI infrastructure businesses as investors scrutinize their financial risks.
The more severe scenario would combine several problems. Disappointing returns could prompt spending cuts, weakening supplier earnings. Lenders could then tighten financing, forcing additional cuts and spreading losses.
What investors should watch
Market watchers suggest the clearest warning would be several major companies simultaneously lowering revenue forecasts or infrastructure spending plans, especially alongside weakening cash flow and more expensive financing.
A recession, new trade restrictions, or a serious security incident could also accelerate a pullback.
A crash isn’t the only possible outcome, however. Stocks could move sideways while earnings gradually catch up with valuations.
For households, the exposure can extend beyond individual AI shares to retirement funds holding major technology companies. The central distinction is that AI’s success as a technology would not guarantee attractive returns on every stock — or at every purchase price.