(Bloomberg) — As the artificial intelligence trade continues to push the stock market to new highs, investors are increasingly questioning whether we are witnessing another financial bubble poised to burst. The situation is not straightforward, particularly when viewed through the lens of historical trends.
The S&P 500 Index (^SPX) surged 16% in 2025, driven largely by AI leaders including Nvidia Corp. (NVDA), Alphabet Inc. (GOOG, GOOGL), Broadcom Inc. (AVGO), and Microsoft Corp. (MSFT). However, there are growing concerns regarding the hundreds of billions of dollars that major tech firms have pledged for AI infrastructure. Combined capital expenditures from Microsoft, Alphabet, Amazon.com Inc. (AMZN), and Meta Platforms Inc. (META) are projected to rise 34% to about $440 billion over the next year, according to data compiled by Bloomberg.
Meanwhile, OpenAI (OPAI.PVT) has committed to spending more than $1 trillion on AI infrastructure, which raises eyebrows given that it is a closely held company without profitability. The cyclical nature of its financial arrangements, where investments and expenditures circulate between OpenAI and several publicly traded tech giants, adds another layer of complexity.
Historically, over-investment has been a recurring theme accompanying transformative technological advancements, as noted by Brian Levitt, chief global market strategist at Invesco. He cited the development of railroads, electricity, and the internet as instances where expectations led to excessive capital inflows. “At some point the infrastructure build may exceed what the economy will need over a short period of time,” he stated. “But that doesn’t mean that the rail tracks weren’t finished or the internet didn’t become a thing, right?”
Despite the S&P 500’s recent performance, with its third consecutive year of double-digit percentage gains, investor anxiety about potential market corrections is palpable. With Nvidia, Microsoft, Alphabet, Amazon.com, Broadcom, and Meta Platforms comprising nearly 30% of the S&P 500, a downturn in AI stocks could significantly impact the index.
Gene Goldman, chief investment officer at Cetera Financial Group, opined that while a bubble typically bursts alongside a bear market, he does not foresee a bear market in the near future. “A bubble likely crashes on a bear market,” he remarked.
To gauge whether the AI-led tech rally has escalated too quickly, one can compare it to previous market bubbles. Research from Bank of America indicates that such bubbles historically last just over two-and-a-half years on average, with an average trough-to-peak gain of 244%. In contrast, the AI-driven rally is currently in its third year, with the S&P 500 climbing 79% since the end of 2022 and the tech-heavy Nasdaq 100 (^NDX) experiencing a 130% increase.
While it is challenging to draw definitive conclusions from these data points, Michael Hartnett from Bank of America cautioned investors against exiting the stock market prematurely. He noted that the final phase of a rally often sees the steepest gains, and missing out could prove costly. To hedge against potential downturns, he suggested diversifying into undervalued sectors such as UK stocks and energy companies.
Currently, the ten largest stocks in the S&P 500 account for approximately 40% of the index, a concentration level not observed since the 1960s. This trend has led some investors to tread carefully. Veteran Wall Street analyst Ed Yardeni remarked that it no longer makes sense to recommend overweighting tech stocks.
Market historians point out that while the current concentration of top stocks appears extreme, similar precedents existed in the 1930s and 1960s. According to Paul Marsh, a professor at the London Business School, in 1900, 63% of US market value was tied to railroad stocks, compared to 37% tied to technology at the end of 2024.
Identifying asset bubbles in real time can be challenging, given that fundamental factors often dominate discussions. Dario Perkins, an economist at TS Lombard, noted the tendency for tech enthusiasts to argue that “it’s different now,” suggesting fundamental valuations may never return to previous norms.
However, some fundamentals remain crucial. Compared to the dot-com bubble, today’s AI leaders exhibit lower debt-to-earnings ratios. Companies such as Nvidia and Meta Platforms are already witnessing robust profit growth attributable to AI, a stark contrast to the situation during the speculative phase of the late 1990s.
Yet concerns surrounding credit risk are emerging; after Oracle Corp. sold $18 billion in bonds on September 24, its stock dropped 5.6% the following day and has since declined 37%. In total, Meta, Alphabet, and Oracle are projected to need to raise $86 billion collectively by 2026, as estimated by Societe Generale.
The S&P 500’s valuation is currently at its highest level aside from the early 2000s, as evidenced by its cyclically adjusted price-to-earnings ratio. This metric, created by economist Robert Shiller, assesses a stock’s price relative to its average inflation-adjusted earnings over the past decade.
Proponents of the market suggest that while valuations are climbing, the rate of increase is significantly slower than during the dot-com era. At one point in 2000, Cisco Systems Inc. was priced at over 200 times its earnings, whereas Nvidia is currently valued at less than 50 times.
As discussions about a potential stock bubble gained momentum through 2023, particularly in November and December, warnings from investor Michael Burry and the Bank of England further fueled the dialogue. A poll by Bank of America revealed that many investors view an AI bubble as the biggest “tail risk” event, indicating a marked shift in sentiment.
This contrasts with the dot-com bubble, where enthusiasm for the internet’s transformative potential was unrestrained. According to Venu Krishna, head of US equity strategy at Barclays, scrutiny of AI investments is increasing, mirroring the doubts that can prevent extreme market fluctuations. “I wouldn’t brush it off, but I would generally think that scrutiny is healthy,” he noted. “In fact, that scrutiny is what will prevent extreme moves like a crash.”
See also
Akakū Announces AI Workshop on Jan. 24 Featuring Suno, Midjourney, and ElevenLabs
Urgent AI Safety Risks Emerge as Breakthroughs Accelerate, Warns UK Research Director
Jayant Chaudhary Urges SPA Graduates to Fuse AI, Ethics, and Creativity for Nation-Building
Amazon Reveals Key Ad Strategies at CES 2026, Prioritizing AI and Live Sports Integration
Illinois Passes Laws to Protect Immigrant Students and Regulate AI Use in Education
