Lessons from the dotcom bubble burst still relevant in AI age 25 yrs later

Exactly 25 years ago, in March 2000, global stock markets hit record highs and then proceeded to collapse, correcting deeply through the next three years. The driver for the bull run was optimism about the internet economy — the so-called dotcom boom. The proximate cause for the correction was investors realising that dotcom stocks were massively overvalued, thus leading to the bubble bursting. The correction continued as events like the terrorist attack on Manhattan’s World Trade Center on September 9, 2001, changed the world. The subsequent invasion of Afghanistan spooked investors and retarded the recovery of the global economy.

 

India’s tech economy matured during the boom-bust cycle with investors waking up to India’s advantages in providing information-technology (IT) services to the world. One trigger was the famous “Y2K” crisis. Older computer systems, which had used only two digits to indicate the year, did not recognise dates like “1 January 2000” and would default back to “1 January 1900” instead. Correcting this required billions of lines of code to be rewritten. Indian IT service providers could deploy a massive workforce of engineers and coders to do this at costs which were a fraction of those incurred by developed-country service providers. These Indian coders could also do more sophisticated tasks. Sorting out Y2K gave Indian IT service providers a proverbial foot in the global door, and as they exploited this, “Bangalored” entered the global business lexicon. Indian IT companies such as Infosys, Wipro, and HCL saw growth at warp speed and valuations zoomed. Many Indian companies issued American depositary receipts — the first time Indian companies were being listed en masse on overseas exchanges. India’s IT sector also became the target of “pump and dump” schemes and outright fraud. A cartel of traders led by Ketan Parekh talked up IT stocks to unrealistic levels. Questionable promoters realised that any company with words like “digital”, “infosystem”, “software” or “tech” in the name would receive fantastic valuations, regardless of fundamentals. These started to unravel in March-April 2000.

 

When financial bubbles inflate, valuations rise to unrealistic levels. When they burst, corrections can also be far deeper than justified by the fundamentals. This happened across the globe with the tech sector. It’s also worth noting that the dotcom bubble both inflated and collapsed long before the promise of the new internet economy translated into concrete profits. Notably, in 2025, only two of the best-performing tech stocks in the world were listed in 2000 — Apple and Microsoft.  Alphabet, Facebook, Netflix, Nvidia — all listed after the bust. Investors were right in that the new technology of the internet did eventually revolutionise the global economy. They were wrong because they were overoptimistic about time to maturity and most of them lost money because their timing was bad. Hundreds of startups raised money through venture-capital and private-equity firms and went public — and most of them died. Remember Yahoo and Netscape? Or Pentafour Software? These were all darlings of investors, but they eventually went bust.

 

In 2024, the world experienced a similar enthusiasm with the promise of artificial intelligence (AI) translating into massive valuations for any AI-related play. Then DeepSeek happened — indicating that massive investment may not be required to build efficient AI agents. Many commentators have compared the AI boom and investment commitment by large companies to the tech boom and bust. It would be worth watching if investors remember market lessons 25 years after the dotcom bust.