Is there an AI bubble and will it burst?

As stocks like Nvidia have risen over 200 per cent in the last year, some are worried an AI bubble is emerging. Jess Jones and Elliot Gulliver-Needham take a closer look

It is hard to escape the hype around artificial intelligence (AI). 

Politicians are convinced it will boost productivity and economic growth, bosses are shoehorning in as many mentions as possible into their earnings calls, and investors, big and small, are wondering how much money they can make off the back of the boom.  

If you had bought £500 worth of shares in chip designer Nvidia – arguably the face of the AI boom – a year and a half ago and sold it today, you would walk away over £3000 richer. Its share price has increased over 200 per cent over the past year.

But, from an investment perspective, some are worried that the good times aren’t going to last forever. 

According to a poll of fund managers last month by the Bank of America, around 40 per cent said that AI-related stocks are a bubble.

Jeremy Grantham, the investor known for predicting the dotcom crash in 2000, has said he believes that AI is a bubble that “will at least temporarily deflate”, if not totally burst.

So, is this an investment bubble that is about to burst or will this technological revolution continue to power strong returns for investors and shareholders? 

Dotcom bubble 2.0?

The benefactors of the AI boom can be split into three different areas. 

These are engagers, who are primarily focused on creating AI-powered products, the enablers, who produce products that enable AI, such as semiconductors, and the enhancers, who are looking to increase the quality of their services through AI, such as Microsoft’s Copilot. 

Stephen Yiu, manager of the Blue Whale Growth fund, stated that Nvidia had become synonymous with AI because the firm is pretty much the only designer of high-powered graphics processing units, or GPUs, which are required to deliver the results to AI queries in real time.

Yiu is incredibly bullish on Nvidia, and has been for some time, stating that he expected the firm’s market cap to surpass that of Apple and Microsoft, valued at $3.2 trillion and $2.6 trillion respectively, to make it the biggest company in the world in the not-too-distant future.

While the incredible growth of Nvidia and other similar companies has been compared to the dotcom bubble, many are eager to point out the differences between the two.

Martin Frandsen, global equities portfolio manager at Principal Asset Management, told CityAM that during the dotcom bubble, “companies that weren’t generating any actual earnings were rewarded with staggering market valuations”.

This is not the case today, he said, where companies such as Nvidia have seen “rapid earnings growth”.

The leading seven tech stocks back then traded on a price to forecast earnings ratio of 52 times in 2000, more than double the Magnificent Seven’s 25 and much higher than Nvidia’s 38.5.

Yiu agreed, adding that while many of the winners of the dotcom bubble had been new firms with poor business models, companies that are currently booming like Nvidia and Microsoft are all “established businesses with diverse revenue streams”.

“Microsoft and Google will not disappear overnight, and Nvidia will still be the supplier of the processors that power their systems,” Yiu told CityAM

Kathleen Brooks, a research director at trading platform XTB, also doesn’t see any parallels between the dotcom bubble and a potential AI bubble because most of the top AI-related companies have strong balance sheets.

“They’re awash with cash and that was very different to what we saw in the 2000s whereby all of a sudden the tide went out and that was it for a lot of tech companies,” she explained.

Although a lot of the major AI-related firms, such as Nvidia and Microsoft, have a high price to earnings ratio, these are justified by their strong earnings and global outlook of their businesses, Brooks said.

“They’re as important as Henry Ford was to the motorcar.”

Kathleen Brooks, a research director at trading platform XTB

Brooks sees sustainable growth for Nvidia going forwards now the company has cemented itself as one of the main designers of chips that allows AI to function, holding around 80 per cent of the global market share of GPUs. 

“They’re as important as Henry Ford was to the motorcar,” she told CityAM 

This could change slightly in the longer term, however, as some of Nvidia’s biggest customers, such as Amazon and Google, attempt to build their own chips, she said, although she noted that this was not quite going according to plan for them right now.

The shakeout

However, there might be some companies riding on the coattails of Nvidia’s success. 

Neil Campling, manager of hedge fund Chameleon Global Capital Management, said there were some AI companies that “have valuations that reflect near utopia”.

He pointed to companies like enterprise AI application software company C3.ai, which is valued at just over $3bn despite losing $2.45 per share, and corporate AI software firm BigBear.ai, which is valued at $467m even though it generated losses of $60m last year and isn’t expected to break even in the next five years. 

Palantir has been another target for AI sceptics, with its big data analytics business quickly incorporating the AI hype into its business model. The company’s stock has shot up 170 per cent in the last year as a result, and is now trading at 66 times its expected yearly earnings.

“There are going to be many use cases that are completely pointless using this technology that people just want to be part of the hype cycle. And so from that perspective, I think that there’s going to be a lot of experimentation, a lot of bullshit companies flying around,” Lewis Liu, co-founder and chief executive of Eigen Technologies, said.

“There is a possibility that smaller, lower quality companies in the AI space that are trading purely on exuberance will disappear, much like the smaller companies of the dotcom bubble.”

Stephen Yiu, manager of the Blue Whale Growth fund

Yiu agreed. “There is a possibility that smaller, lower quality companies in the AI space that are trading purely on exuberance will disappear, much like the smaller companies of the dotcom bubble.”

Few, if any, of the younger companies, such as OpenAI, are currently running on a positive gross margin, even if they are showing top line growth.

“We may end up in an Uber situation,” said Liu, “where right now [OpenAI] is subsidised by venture capital and Microsoft, but the original ambition is tempered down significantly.”

However, unlike the dotcom bubble, almost all of the key AI startups are privately traded, meaning they are less vulnerable to the whims of the market and to things immediately bubbling up and popping.

Some of the big names of the AI boom are increasingly self-aware of the investor excitement that surrounds them. A recent report by The Information revealed that executives at big tech firms including Amazon, Microsoft and Google are quietly dialling down generative AI expectations internally, particularly in sales, amid concerns that the technology can not live up to the hype.

Whether or not there is an AI ‘bubble’, there is likely to be a “shakeout”, argued Liu.

“That means that we will see a big drop in valuation in some of these AI darlings and we may see some of them go bust or just get acquired for scraps,” he explained. “I think we’re going to see a revolution, but there is going to be a shakeout at the same time.”