November 22, 2025
What goes up has to come down is something that is frequently used with reference to the movements in the equity market. It also refers to the literal force of gravity and also suggests a predictable decline in what has achieved fame & glory. In the financial lingo it also finds concurrence with the term bubble. We hear this word more frequently nowadays thanks to the optimistic outlook in a sector called AI. p.s. AI refers to Artificial Intelligence and not Air India as I did some time ago!
Unfortunately, such bubble-like situations find validation only when they burst. Now for us to analyze whether the AI sector is a potential bubble, one needs to deep dive into its exponential rise and its share in American equities. As per available data from JP Morgan, 30 AI linked companies have a market share of 44% of the S&P 500 index. Just five companies, namely, Nvidia, Microsoft, Apple, Alphabet & Amazon have a representation of 29% in the same index. So, in terms of concentration risk, we clearly have a tick mark.
When we look at valuations, we have two startups based on the AI flavour, Nano Nuclear Energy, being valued at 2.30 billion USD despite having no revenue or even a license to operate. The other being Fermi, valued at 14.80 billion USD. Nvidia is currently trading at a P/E ratio of 54 times (trailing) and the historical range being 46 to 84. Yes, there is a wave of optimism in spite of stretched valuations and startups gaining unimaginable valuations just with the AI tag but nil bottom line. How different is this situation from the tech bubble of 1999-2001. There were a bunch of companies which sounded like Infosys when pronounced but with a different spelling. This was to ride the wave of the standards created by Infosys and to get noticed by investors.
If someone had to go by the legendary fund manager, Michael Burry’s conviction, we should be exiting AI companies like Palantir & also Nvidia. By the way, Mr. Burry had pocketed 100 million USD and his fund delivered roughly 700 million USD for its investors, just by shorting the housing mortgage companies in 2008, when the whole world including the top Banks were investing more into it. For better insights, I recommend you to watch the movie, Big Short.
If it is AI that we are contemplating on whether it is a bubble or not, you will be surprised to know that it was Tulips in 1630 in the Netherlands. It was a manic phase where people kept betting on the further price rise in Tulips, yes tulip a pretty flower. It reached its zenith when the value of a few rare tulip bulbs surpassed the cost of a luxury house. Then one fine day, the tulip prices just crashed, to the tune of 90%. Any guesses on how? The buyers just did not show up at the auction. Same is the case with AI today, as long as there are presentations from the management on projecting its next level of growth and there are huge funds willing to buy the story, prices of AI companies will keep rising. But if the investors decide to withdraw 50% of their holdings on a single day, anything can collapse. So long as optimism is high, prices tend to go up.
How can one forget the Japanese experience of the 1980s? Somewhere in 1986, the Yen appreciated by 45% and Japan was staring at a recession. Currency appreciation is harmful for export-oriented countries. The Bank of Japan (BOJ) had to inject liquidity to revive the sinking economy. This excess money made way into the stock market and also jacked up real estate prices. The P/E ratio of the Japanese index, Nikkei 225, touched 80 and the land prices hit the roof. BOJ realised its mistake and started to increase its rates only to reduce the spending power of the people and subsequently the markets crashed close to 15% in a day and went on to lose 80% of its value for the next 18 years. Yes, a prolonged correction for 18 years. This relentless crash bottomed out only in 2009. It managed to reclaim its previous peak of 1989 only in the year of 2024!
If it was tulips in the 1600s then, tech in the late 1990s, housing in 2008 we now have AI. Whether it is a bubble, we need to wait for it to burst.