What if the AI stockmarket blows up?(economist.com)
economist.com
What if the AI stockmarket blows up?
https://www.economist.com/finance-and-economics/2025/09/07/what-if-the-ai-stockmarket-blows-up
3 comments
https://archive.is/wYpuQ
Practically none of the AI companies are traded on the stockmarket, so this seems more like a firecracker going off at most.
>"Tech enthusiasm always runs ahead of tech realities," according to Michael Parekh, a former analyst at Goldman Sachs
This is more true now than ever. The fact that companies with thin floats trade a few shares at high prices (which is then multiplied to the entire float to generate a "billion dollar" market cap) is a well rehersed game these days, in both private and public markets. Even retail investors generally understand this (not always). It's "priced in" to some extent.
>One lesson from history is that, when tech bubbles burst, leading firms often give way to upstarts. “The biggest and most successful lighting companies all experienced a change of control when cashflow became an issue,” wrote Alasdair Nairn in “Engines That Move Markets”, a history of the late 19th century. Many firms that dominated the early days of railways, the telegraph and the telephone were also quickly supplanted. Who now remembers Vulcatron, from America’s electronics bubble of the 1960s, or Corning, a household name during the dotcom boom? It will be a miracle if, in a decade or so, all of the “magnificent seven” listed tech firms, and the biggest AI startups, still exist.
I have a somewhat hard time following the logic here. History rhymes but it isn't necessarily a blueprint. If the "AI bubble" bursts to the extent that Meta/Google/Amazon lose their position, that means it's a fairly catastraughic scenario, considering they have fortress balance sheets and pseudo-monopolies that are being actively defended by their government (US red lines vs digital services taxes in trade negotiations). It would obviously mean that their AI datacenter investments are kaput, but if that's the case, then the Coreweaves and Nebiuses and definitely the VC sky high valuations are coming down 10x if not 100x as hard.
Therefore the conclusion here is that the upstarts are lurking in the metaphorical startup-garages, while the bubble inflates outside. Maybe true, but it sure seems like recency bias anchored to the DotCom narrative, which may or may not play out again. I'd personally argue that the 2020 unwind of unprofitable tech with 10 billion dollar smart-treadmill companies, 100 (sorry, 160) billion dollar video chat software, 10B space tourism, and $90 billion dollar EV companies (lucid motors), was this generation's blindingly obvious DotCom bust.
>Many firms that dominated the early days of railways, the telegraph and the telephone were also quickly supplanted. Who now remembers Vulcatron, from America’s electronics bubble of the 1960s, or Corning, a household name during the dotcom boom? It will be a miracle if, in a decade or so, all of the “magnificent seven” listed tech firms, and the biggest AI startups, still exist.
This is what seems like the biggest grey swan, but the AI tangent is almost secondary to the driving factor, which is index concentration and the reflexive feedback loop which drives more and more passive dollars to the biggest companies. The historical parallel here was oddly not even named in the article, which is the Nifty Fifty bubble in the mega-conglomerates in the 70s.
What happened is their internal flywheel became a drag instead of a boost, and this was a natural result of a reversal of the forces that drove them to the top (high equity valuation, high growth/economic conditions they could harness). Some of the new financial engineering in the AI world (debt flywheels securitized by GPUs and future locked in revenue) has some obvious risks, but when it comes to the hyperscalers like Google and Amazon, we have to be aware that they have enormously resilient balance sheets. The likely outcome is a big valuation re-rate. Say the 27x PE comapnies move to 15x.
>Should Digital God fail to arrive, the fall will be brutal.
I'm not entirely convinced? The case they laid out seems to be potentially contained explosions. Pensions could be hurt. Family offices, definitely. It wouldn't be fun, but I'm not sure it would equal even the DotCom bust (speaking as someone who saw the financial ruin of destroyed Dotcom companies/hero to zero teleportation firsthand). I think the stock concentration bubble is the more obvious risk, but if that goes down, that portends wider economic pain that arguably overshadows the AI downturn. Indeed, the AI narrative may simply lose focus, and continue building without hype (maybe it will even be treated with disdain in the investing space), while a wider economic slowdown emerges.
This is more true now than ever. The fact that companies with thin floats trade a few shares at high prices (which is then multiplied to the entire float to generate a "billion dollar" market cap) is a well rehersed game these days, in both private and public markets. Even retail investors generally understand this (not always). It's "priced in" to some extent.
>One lesson from history is that, when tech bubbles burst, leading firms often give way to upstarts. “The biggest and most successful lighting companies all experienced a change of control when cashflow became an issue,” wrote Alasdair Nairn in “Engines That Move Markets”, a history of the late 19th century. Many firms that dominated the early days of railways, the telegraph and the telephone were also quickly supplanted. Who now remembers Vulcatron, from America’s electronics bubble of the 1960s, or Corning, a household name during the dotcom boom? It will be a miracle if, in a decade or so, all of the “magnificent seven” listed tech firms, and the biggest AI startups, still exist.
I have a somewhat hard time following the logic here. History rhymes but it isn't necessarily a blueprint. If the "AI bubble" bursts to the extent that Meta/Google/Amazon lose their position, that means it's a fairly catastraughic scenario, considering they have fortress balance sheets and pseudo-monopolies that are being actively defended by their government (US red lines vs digital services taxes in trade negotiations). It would obviously mean that their AI datacenter investments are kaput, but if that's the case, then the Coreweaves and Nebiuses and definitely the VC sky high valuations are coming down 10x if not 100x as hard.
Therefore the conclusion here is that the upstarts are lurking in the metaphorical startup-garages, while the bubble inflates outside. Maybe true, but it sure seems like recency bias anchored to the DotCom narrative, which may or may not play out again. I'd personally argue that the 2020 unwind of unprofitable tech with 10 billion dollar smart-treadmill companies, 100 (sorry, 160) billion dollar video chat software, 10B space tourism, and $90 billion dollar EV companies (lucid motors), was this generation's blindingly obvious DotCom bust.
>Many firms that dominated the early days of railways, the telegraph and the telephone were also quickly supplanted. Who now remembers Vulcatron, from America’s electronics bubble of the 1960s, or Corning, a household name during the dotcom boom? It will be a miracle if, in a decade or so, all of the “magnificent seven” listed tech firms, and the biggest AI startups, still exist.
This is what seems like the biggest grey swan, but the AI tangent is almost secondary to the driving factor, which is index concentration and the reflexive feedback loop which drives more and more passive dollars to the biggest companies. The historical parallel here was oddly not even named in the article, which is the Nifty Fifty bubble in the mega-conglomerates in the 70s.
What happened is their internal flywheel became a drag instead of a boost, and this was a natural result of a reversal of the forces that drove them to the top (high equity valuation, high growth/economic conditions they could harness). Some of the new financial engineering in the AI world (debt flywheels securitized by GPUs and future locked in revenue) has some obvious risks, but when it comes to the hyperscalers like Google and Amazon, we have to be aware that they have enormously resilient balance sheets. The likely outcome is a big valuation re-rate. Say the 27x PE comapnies move to 15x.
>Should Digital God fail to arrive, the fall will be brutal.
I'm not entirely convinced? The case they laid out seems to be potentially contained explosions. Pensions could be hurt. Family offices, definitely. It wouldn't be fun, but I'm not sure it would equal even the DotCom bust (speaking as someone who saw the financial ruin of destroyed Dotcom companies/hero to zero teleportation firsthand). I think the stock concentration bubble is the more obvious risk, but if that goes down, that portends wider economic pain that arguably overshadows the AI downturn. Indeed, the AI narrative may simply lose focus, and continue building without hype (maybe it will even be treated with disdain in the investing space), while a wider economic slowdown emerges.