The central bank of the United Kingdom is worried about an AI bubble burst.
“On a number of measures, equity market valuations appear stretched, particularly for technology companies focused on Artificial Intelligence (AI),” the bank’s financial policy committee said, according to a record of its latest meeting. “This, when combined with increasing concentration within market indices, leaves equity markets particularly exposed should expectations around the impact of AI become less optimistic.”
The Bank also warned that stock market price valuations were comparable to the peak of the dot-com bubble, and the market share of the top five members of the S&P 500 was at its highest concentration in 50 years. Those five companies are, unsurprisingly, AI-focused tech giants Nvidia, Microsoft, Apple, Amazon, and Meta.
All five of these companies are spending eye-watering figures on AI, and the stock market loves it. Microsoft became the second company to ever hit a $4 trillion market valuation earlier this year after posting its largest ever quarterly expenditure forecast. Nvidia, on the other hand, is the first and only company in the world to hit a $4.5 trillion market cap.
“Material bottlenecks to AI progress – from power, data, or commodity supply chains – as well as conceptual breakthroughs which change the anticipated AI infrastructure requirements for the development and utilisation of powerful AI models could also harm valuations, including for companies whose revenue expectations are derived from high levels of anticipated AI infrastructure investment,” the bank said.
Fed researchers issued a similar warning earlier this year. While that alert did not identify an immediate risk of an AI bubble, the researchers pointed out that a risk that comes with building expensive infrastructure too quickly for anticipated demand was that demand might not grow as expected. In that case, it could lead to “disastrous consequences,” the Fed warned, likening it to the railroad over-expansion of the 1800s that led to an economic depression towards the turn of the century.
These top AI companies with high revenue expectations are also heavily reliant on each other financially, increasing worries of a cascade effect if a bubble bursts. AI companies ink multibillion-dollar deals with each other over and over again, injecting more money into the system and ballooning stock valuations with each deal.
While that’s happening, some experts are admitting overvaluation.
Apollo Global Management’s chief economist Torsten Slok said in July that the AI bubble of today is actually worse than the 1999 dot-com bubble. OpenAI CEO Sam Altman also admitted in August that he thinks investors are “over-excited about AI.”
The main downside risks of AI overvaluation, according to the bank, also include disappointing AI capability or adoption progress.
A recent MIT report found that despite the major push to adopt AI in the corporate world, fewer than one in ten AI pilot programs actually generated real revenue gains. The report spooked investors enough that AI stocks immediately slid following the headlines in August.
Last month, the Census Bureau showed that the rate of AI adoption by large companies had been declining slightly.
Nonetheless, executives keep assuring investors that AI demand is scaling rapidly as the technology finds its way into more and more areas of life. AI computing demand is up “substantially” in the past six months, according to Nvidia CEO Jensen Huang’s comments on Wednesday.
But if the tech giants are wrong and the Bank of England’s risk scenario does end up being the case, the bank warned that a sudden, sharp correction could occur, “adversely affecting the cost and availability of finance for households and businesses.”
The U.S. has reason to be afraid of this. According to recent reports, the AI spending frenzy is not just propping up the American stock market, but it’s also lifting the real economy. According to Harvard economist James Furman‘s calculations, U.S. GDP growth in the first half of the year was almost entirely driven by investments made in data centers and other information-processing technology.