For now, those productivity gains seem far from materialising, which has led to a lot of babble about a bubble. “Bubbles are among the oldest stories of capitalism,” notes Azeem Azhar, a British tech entrepreneur. “They’re parables of excess, belief and collapse.”
From the Dutch Tulip Mania to the South Sea frenzy of the 1720s, roaring stock market of the 1920s, Japan’s property boom in the 1980s and the US subprime burst and housing crash that triggered the 2008 global financial crisis, he points out that bubbles are really a tale of greed and folly. “Bubbles become stories we tell ourselves about the dangers of optimism.”
Warnings of an imminent AI bubble pop are growing louder by the day. Goldman Sachs CEO David Solomon and e-commerce giant Amazon’s founder Jeff Bezos joined the chorus this past week. “I wouldn’t be surprised if in the next 12 to 24 months, we see a drawdown,” Solomon said, stopping short of calling it a bubble while making it clear that investors may be getting a little too giddy.
Bezos, for his part, painted a picture of an environment in which early-stage ideas are getting funded despite little substance. “This is the kind of industrial bubble,” said Bezos, who had steered Amazon from the verge of bankruptcy in the aftermath of the dotcom bust.
See also: Bank of Singapore uses AI agents to cut source of wealth report time to one hour
Even Sam Altman, CEO of the world’s largest tech start-up OpenAI which created ChatGPT, recently warned that investors were getting way “overexcited” about AI. Last week, a share sale valued OpenAI at US$500 billion ($646 billion), or more than the market cap of oil giant Exxon Mobil and streamer Netflix.
Signs of froth
So, are we really in another tech bubble 25 years after the dotcom bust? Look no further than the AI spending boom for signs of froth. A key sign of froth is tech research firm Gartner expecting AI-related spending globally to surpass US$2 trillion next year — or 2% of global GDP.
See also: The AI-robotics revolution, China-US rivalry and Southeast Asia
The latest bull run began three years ago, when OpenAI unveiled ChatGPT. Having fallen 24% in the first 10 months of 2022, the S&P 500 began a sharp rebound. The index was up 26.3% in 2023 and 25% last year; and has gained 14.5% this year to date, or 15.9%, if you include reinvested dividends.
Tech stocks have understandably done even better. The Nasdaq 100, which includes the top 100 US tech stocks, is up 18.4% this year and 65% higher after the “Liberation Day” tariff lows in April.
Shares of the Magnificent 7 — the top seven US tech firms: Nvidia, Microsoft, Apple, Google’s owner Alphabet, Amazon, Tesla and Meta Platforms — are up 23% this year and have more than doubled since January last year.
Yet, it is probably a stretch to describe the run as a “bubble”. The average US bull market over the century has lasted 59 months, with a return of 178%. The current bull market is in its 35th month with an 83% return.
If history is any guide, the current bull run has some way to go, particularly in the context of the bull market of the 1990s that eventually morphed into a bubble in 2000.
That bull market lasted exactly 10 years from April 1990 to March 2000, with total S&P 500 gains of about 420%.
Are warning bells in order then? In late 1999, at the height of the dotcom bubble, billionaire investor Warren Buffett described what happens when a plain-vanilla bull market is on the verge of morphing into a real bubble.
Sink your teeth into in-depth insights from our contributors, and dive into financial and economic trends
“Once a bull market gets underway, and you reach the point where everybody has made money no matter what system he or she followed, a crowd is attracted into the game that is responding not to interest rates and profits but simply to the fact that it seems a mistake to be out of stocks,” Buffett said just months before the bubble popped in March 2000.
Clearly, by the Sage of Omaha’s measure, we are not quite there yet. The 95-year-old CEO of Berkshire Hathaway, who retires at the end of December, believes bull markets end abruptly when the rally is powered by momentum, quite aside from stretched valuations and irrational exuberance.
While valuations might be a tad stretched, particularly in parts of tech, Wall Street pundits point out that we are still not seeing anywhere near the sort of momentum or irrational exuberance that preceded the March 2000 bubble burst.
Between September 1998 and January 1999, in the aftermath of the Asian financial crisis, the Russian currency and banking crisis, and the collapse of the hedge fund Long-Term Capital Management, where two Nobel laureates Robert Merton and Myron Scholes worked, the US Federal Reserve began slashing interest rates.
That avalanche of liquidity helped fuel the dotcom bull run. That bubble burst in March 2000 only after the Fed had raised interest rates by 175 basis points within 11 months, the third fastest rate-hiking cycle on record.
Sure, there was overvaluation, exuberance and momentum buying as everyone chased their favourite internet stock — from Pets.com to Webvan, which lacked a viable business model, a decent revenue stream, a customer base or indeed even basic technology, but Fed hikes were the final nail that popped the bubble.
On Oct 2, shares of AI icon Nvidia Corp touched an all-time high of US$191, giving the chip behemoth a market cap of US$4.64 trillion — 50% more than the combined market value of US$3.06 trillion for all listed companies in Southeast Asia.
The S&P 500, the main US benchmark, touched at an all-time high of 6,754 this past week. At 6,754, the S&P 500 is trading at 24.9 times this year’s earnings. Consensus S&P 500 earnings forecasts for 2026 are around US$305 per share, as such the market is trading at around 22 times next year’s earnings.
The earnings growth forecast for next year is around 12%. The bull market might look a tad pricey in its fourth year, but as the AI-powered growth continues, some of the froth may be justified.
Liquidity avalanche
For one thing, a flood of liquidity is about to become an avalanche as the US Fed resumes its interest rate cutting cycle this month. With the US government shutdown after President Donald Trump’s Republican Party failed to agree with the opposition Democrats on a spending bill, the Fed is now more likely than not to keep cutting rates because of its dual mandate of tackling unemployment and inflation.
The Trump administration, which has drastically slashed the size of the government since it assumed office in January, has hinted that it sees the shutdown as an opportunity to cut more employees, which will likely force the Fed to do more than it wants to despite signs that inflation may be creeping up again.
Will there be demand for Nvidia’s graphic processing units (GPUs) that are used for AI training, or Broadcom Inc’s customised AI chips, or auxiliary processing units (XPUs)? Demand for AI compute is expected to outstrip supply again next year and the year after.
Citigroup last week raised its total capex forecast for giant cloud service providers Amazon Web Services, Microsoft Azure, Google Cloud and Oracle to US$490 billion for 2026, and their total spending between 2026 and 2029 to a whopping US$2.8 trillion. It now expects Microsoft Azure to spend US$127 billion in 2026, or up 43% over the current year.
The bank also raised its estimates for Oracle’s capex to US$57 billion during the next fiscal year beginning June 2026.
Everyone, everywhere, seems to want a piece of the action. Investors believe the market is headed higher despite the fact that valuations are elevated and the bull run is being driven by the Magnificent 7 plus custom AI chip maker Broadcom, and software firms Oracle and Palantir, that are now formidable AI players.
The market’s run in recent weeks has begun to draw in novice investors eager to profit from a transformative new technology. Many have likened it to the meme stock mania of 2021 when young retail investors took a cue from Reddit’s wallstreetbets community and helped catapult massively shorted stocks like GameStop to new heights.
To others, it is all about investors’ FOMO — the fear of missing out. For their part, Wall Street investment banks, while warning of “overvaluation” in some parts of tech, are generally of the view that large-cap US tech stocks are still currently in a secular bull run driven by monetisation opportunities around generative AI.
Nvidia, Microsoft, Alphabet and Meta have much better fundamentals than Cisco Systems Inc and Amazon had at the height of the dotcom bubble.
“Today’s market, particularly the tech sector, exhibits dotcom-era overvaluation, with lofty multiples, slower earnings growth and a weaker macroeconomic backdrop,” GQG Partners, a Florida-based investment firm headed by billionaire Indian investor Rajiv Jain, said in a report titled Dotcom on Steroids last month. “We believe today’s technology sector no longer represents forward-looking quality due to decelerating revenue growth, collapsing free cash flow and increasing competition.”
Since the 2008 financial crisis, the GQG report pointed out, the US tech sector has been the standout investment trade, defying the concerns of value investors of overly steep valuations.
“While many initially underestimated the business quality, growth runway and long-term earnings power of big tech, these companies — led by visionary founders — evolved into monopolistic giants, delivering fast growth and robust profit margins,” it noted. “In a growthstarved, zero-interest-rate world that continuously drove capital toward secular growing compounders, this was the perfect set-up for massive outperformance.”
Now, however, the sector stands at a significant inflection point, “with investors seemingly making a one-way bet on the Al mania while appearing to ignore alarming fundamental issues”, GQG said.
The momentum in growth-oriented segments of the market — including big tech and firms tied to the Al infrastructure buildout — could reverse at any moment, the report argued.
“Frothy, but not a bubble, yet,” says Venu Krishna, a strategist for Barclays in New York. A bubble is least likely to pop when everyone is babbling about it. So, what should rational investors do as the AI bubble inflates? “
When I see a bubble forming, I rush in to buy, adding fuel to fire,” legendary hedge fund manager George Soros once said. “That’s not irrational,” he argued.
Like the internet 25 years ago, AI is a truly disruptive technology. There will be pullbacks, just like in every bull run. Valuations “remain below the peak [they] reached in the dotcom bubble when investors last sought to crystallise the future benefits of a transformative technology”, notes John Higgins at research firm Capital Economics in London. “What’s more, unlike in the final stages of that bubble, the Fed doesn’t seem set to spoil the party.”
Eventually, the bubble will pop, but before then expect “the US stock market to charge ahead over the rest of this year and next, with other tech-heavy markets — such as China’s — not too far behind”, says Higgins.
Assif Shameen is a technology and business writer based in North America
