I disagree. I find the relatively recent 2020 bubble to be closer to the DotCom bubble.
100 billion dollar video conferencing tools, 50 billion dollar treadmills, 20 billion dollar EV startups that are vaporware, and space tourism companies valued over 10 billion that are hot air.
In the public equity space, much of the AI market cap is contained in megacap tech companies which have fortress balance sheets and strong cashflows to support the capex and AI initiatives. That begs the question "are they overvalued?", and well, yes, they could be, but that would move the needle closer to the Nifty 50 historical parallel than the DotCom bubble. The DotCom bubble at its zenith was closer to the massive 2020 "profitless tech" bubble (or duration bubble, to be more general), in that it devolved into pure fantasy. Indeed, it rewarded fantasy, and punished realism and anything that brought the narrative down to earth (like profit at any expense to future growth).
It's very strange to me how the massive 2020 bubble is glossed over. That was the most DotCom-esque speculation I've seen since, well, the DotCom bubble, and it's something I definitely had exposure to due to doing relative value SPAC warrant trading over that timeframe on dozens/hundreds of those names. I'm trading these AI hype names too, but today isn't really close to DotCom or 2020 when it comes to tone. Yes retail participation is high, but it's mostly in names like Palantir and not a full on delusional speculative mania. DotCom was also full of outright scams. 2020 was full of outright near-criminal level scams as well, with projections fabricated by hype-men to an insane degree (see image below to sum it up), eventually requiring the SEC to step in and shut down 90% of the SPAC IPO machine with threats of criminal proceedings. Today's AI startups feel very "modern venture capital" in comparison, and at least the typical SaaS metrics like ARR aren't totally ignored in favor of fantasy narratives (well, maybe mostly, but not close to full on DotCom mania levels).
This was an actual slide from SoftBank's investor deck during the 2020 DotCom 2.0 bubble (perhaps the best slide of all time), explaining how they were going to weather the storm with their massive hoard of speculative investments: https://i.ibb.co/MknYdQdY/valley.jpg
Make no mistake, the forward returns of these rich AI names are probably quite terrible, I just honestly think today's market is more like Nifty50 than DotCom. And while 2020/DotCom style unwinding mainly hurts the speculators, the index concentration unwinds - ex: NVIDIA down 70%, PLTR down 90%, etc - can have more of an impact on real 401ks than Virgin Galactic and Nikola going to zero, because long market cap weighted indexes is the ultimate crowded trade.)
Howard Marks just published a memo about this:
https://www.oaktreecapital.com/insights/memo/the-calculus-of...
Personally I think AI is underhyped.
I disagree. I find the relatively recent 2020 bubble to be closer to the DotCom bubble.
100 billion dollar video conferencing tools, 50 billion dollar treadmills, 20 billion dollar EV startups that are vaporware, and space tourism companies valued over 10 billion that are hot air.
In the public equity space, much of the AI market cap is contained in megacap tech companies which have fortress balance sheets and strong cashflows to support the capex and AI initiatives. That begs the question "are they overvalued?", and well, yes, they could be, but that would move the needle closer to the Nifty 50 historical parallel than the DotCom bubble. The DotCom bubble at its zenith was closer to the massive 2020 "profitless tech" bubble (or duration bubble, to be more general), in that it devolved into pure fantasy. Indeed, it rewarded fantasy, and punished realism and anything that brought the narrative down to earth (like profit at any expense to future growth).
https://i.ibb.co/d4c7kD9S/bigmoo.webp (and we saw how that chart ended...)
It's very strange to me how the massive 2020 bubble is glossed over. That was the most DotCom-esque speculation I've seen since, well, the DotCom bubble, and it's something I definitely had exposure to due to doing relative value SPAC warrant trading over that timeframe on dozens/hundreds of those names. I'm trading these AI hype names too, but today isn't really close to DotCom or 2020 when it comes to tone. Yes retail participation is high, but it's mostly in names like Palantir and not a full on delusional speculative mania. DotCom was also full of outright scams. 2020 was full of outright near-criminal level scams as well, with projections fabricated by hype-men to an insane degree (see image below to sum it up), eventually requiring the SEC to step in and shut down 90% of the SPAC IPO machine with threats of criminal proceedings. Today's AI startups feel very "modern venture capital" in comparison, and at least the typical SaaS metrics like ARR aren't totally ignored in favor of fantasy narratives (well, maybe mostly, but not close to full on DotCom mania levels).
This was an actual slide from SoftBank's investor deck during the 2020 DotCom 2.0 bubble (perhaps the best slide of all time), explaining how they were going to weather the storm with their massive hoard of speculative investments: https://i.ibb.co/MknYdQdY/valley.jpg
Make no mistake, the forward returns of these rich AI names are probably quite terrible, I just honestly think today's market is more like Nifty50 than DotCom. And while 2020/DotCom style unwinding mainly hurts the speculators, the index concentration unwinds - ex: NVIDIA down 70%, PLTR down 90%, etc - can have more of an impact on real 401ks than Virgin Galactic and Nikola going to zero, because long market cap weighted indexes is the ultimate crowded trade.)