Khaberni - As the stock market continues to boom thanks to artificial intelligence, investors are increasingly wondering if we are witnessing another financial bubble that will inevitably burst.
The answer to the artificial intelligence bubble is not that simple, at least according to history.
The S&P 500 index jumped 16% in the year 2025, with prominent AI companies such as Nvidia, Alphabet, Broadcom, and Microsoft having the lion's share of this rise. However, at the same time, concerns are growing about the hundreds of billions of dollars that tech giants have pledged to spend on AI infrastructure, according to a Bloomberg report seen by "Arabiya Business".
According to data collected by Bloomberg, the capital expenditures of Microsoft, Alphabet, Amazon, and Meta Platforms will increase by 34% to reach about 440 billion dollars collectively next year.
Meanwhile, OpenAI has committed to spending more than a trillion dollars on AI infrastructure, an astounding amount for a private, nonprofit company. However, perhaps more concerning is the circular nature of many of its deals, where investments and spending are exchanged between "OpenAI" and a few stock market-listed tech giants.
Throughout history, over-investment has been a common feature when technological progress is poised to transform society, according to Brian Levitt, Global Market Strategist at Invesco for investment management, referring to the development of railroads, electricity, and the internet. It may not be different this time.
Levitt said: "At some point, infrastructure build-up may exceed what the economy will need for a short period... but this doesn’t mean that the railroads were not completed or that the internet did not become a reality, right?"
However, with stock valuations rising gradually, and the S&P 500 marking its third consecutive year of two-digit gains, it makes sense for investors to grow concerned about the size of remaining gains and market value that might be lost if AI fails to meet expectations.
Nvidia, Microsoft, Amazon, Alphabet, Meta, and Broadcom comprise about 30% of the S&P 500, so any sharp decline in AI stocks would heavily impact the index.
Gene Goldman, Chief Investment Officer at Cetera Financial Group, who does not believe AI stocks are in a bubble, says: "It's more likely that the bubble will burst in a bear market. We just do not expect a bear market anytime soon."
Here is how the current AI boom compares to past market bubbles.
Velocity and Duration
One simple way to evaluate whether the AI-driven technological surge has gone too far or too quickly is to compare it with past periods of rise. Looking at 10 stock bubbles from around the world since 1900, these typically lasted more than two and a half years on average, with gains reaching 244% from lowest to highest levels, according to research conducted by Michael Hartnett, a strategist at Bank of America.
Compared to that, the current AI-driven rise is in its third year, with the S&P 500 index up 79% since the end of 2022, and the Nasdaq 100 index, which is heavily weighted towards technology stocks, up 130%.
Difficulty interpreting data suggests caution for investors eager to sell their stocks even if they believe they are in a bubble, as the last phase of the rise is typically the most intense, and missing this opportunity can be costly. Hartnett suggests that one hedging strategy is to buy stocks of companies with low valuations, such as British and energy companies.
Focus
The top 10 stocks currently make up about 40% of the S&P 500 index, a level of concentration not seen since the 1960s. This has led some investors, including Ed Yardeni, a veteran Wall Street researcher who stated in December that it no longer made sense to recommend increasing the weight of technology stocks in their portfolios.
Market historians argue that, although this concentration may seem excessive compared to recent memory, there are precedents. In the 1930s and 1960s, the proportion of major company stocks in the American market was similar, according to Paul Marsh, a professor at London Business School, who studied global asset returns over the past 125 years.
Marsh stated that back in 1900, 63% of the American market value was tied to railroad companies, compared with 37% linked to technology at the end of 2024.
Fundamentals
Detecting asset bubbles in real time tends to be far more difficult than identifying them after the fact, because the fundamentals are usually the focus of discussion, and the metrics focused on by investors may change, according to economist Dario Perkins from TS Lombard.
Perkins said: "It's easy for tech enthusiasts to claim that 'it's different this time', and that fundamental valuations will never revert to what they once were."
But some fundamentals always matter. For example, compared to the internet bubble, today's major AI companies have lower debt-to-earnings ratios than companies like WorldCom. Moreover, companies like Nvidia and Meta Platforms are already reporting strong profit growth thanks to AI, which was not necessarily the case during the speculative era 25 years ago.
This credit risk probability in AI trading makes some investors nervous. For instance, after Oracle sold bonds worth 18 billion dollars on September 24, its stock fell by 5.6% the next day, and since then, it has dropped by 37%. Meta, Alphabet, and Oracle will need to collectively raise 86 billion dollars in 2026 alone, according to estimates by Société Générale.
Valuations
The S&P 500 index currently has the highest valuation ever, except for the early 2000s, according to the cyclically adjusted price-to-earnings ratio, a measure devised by economist Robert Shiller, where the stock price is divided by the inflation-adjusted average of its earnings over the past ten years.
Optimistic investors argue that while market valuations are rising because of technology, the pace of increase is much slower compared to the dot-com boom. At one point in 2000, the price of a Cisco Systems share was more than 200 times its earnings over the previous 12 months, while Nvidia's stock today is less than 50 times that.
Richard Claude, a fund manager at Janus Henderson Asset Management, believes that stock prices detach from earnings growth in an environment where valuations are not debated, and he says: "We do not see that happening as of now."
Investor Scrutiny
Debates about a potential stock market bubble have been spreading throughout the year, but intensified notably in November and December amid warnings from investor Michael Burry and the Bank of England. According to data gathered by Bloomberg, the term "AI bubble" appeared in over 12 thousand news articles in November alone, roughly equivalent to the total from the previous ten months.
A poll conducted by Bank of America in December showed that investors saw the AI bubble as the greatest "long-shot risk". More than half of the poll participants said that stocks of the seven big tech companies were the most traded in Wall Street.
This contrasts with the web bubble, when there was "full excitement around the internet revolutionizing everything", according to Venu Krishna, head of US equity strategy at Barclays. As debt issuance rises, questions grow about whether investments in AI will bear fruit or not.
