top of page
Search

Is the AL rout like the Dot.com bubble i 2002? We going down?

snitzoid
12 minutes ago
4 min read

Prior to the Dot.com bubble, the Forward PE for the SP 500 was a frothy 26. In November before the current market correct the number stood at 18. Does that mean we're out of the woods? Suspect not.


Are we in for the down escalator that occurred in 2003/4? Don't think so. It's not unlikely that resolving the Iran crisis and Ukraine may trigger a sharp rebound in the next 12-18 months.


Unless I'm wrong.


Has the AI Bubble Already Burst?

There are remarkable similarities in stock price declines during the dot-com bubble and today


By James Mackintosh, WSJ

Nov. 22, 2026 5:30 am ET


History may not repeat itself exactly, but the AI selloff ravaging the stock market in recent weeks is remarkably like the bursting of the dot-com bubble. As investors continue to worry about whether there’s a bubble in AI, a valid question to ask is whether there was, and has it already popped?


Consider some of the big technology names from the dot-com era that are still around. Here are the drops from their high to the end of the year in 2000 and in 2026:


Intel -60% in 2000 vs. -64% in 2026

Oracle -37% vs. -60%

Cisco -52% vs. -30%

Micron -63% vs. -53%

Qualcomm -54% vs. -42%

Seagate Technology -37% vs. -60% (It merged and was spun out again.)

Nvidia -62% vs. -48%


Some are almost identical, some worse, some better. But just from these it would be hard to say one was obviously a burst bubble and the other not.


Back in 2000 the falls were widespread enough to drag down the S&P 500 by almost 15% from its high to the end of the year, despite a huge rotation from tech into cheap and defensive stocks (utilities and consumer staples both rose more than 40%!).


This year the selloff has been concentrated in AI-related stocks, and again there’s been a big rotation—including healthcare up 24% from late April—but still the S&P is down 12% from its intraday high.



A big difference, so far, is that the dot-com bust led to a recession, which contributed to weakness in sectors otherwise unaffected and spurred another deep selloff in 2001. The unknown now is whether the selloff continues, as it did then.


The market’s failure to rally after the successful IPO of Anthropic on Thursday isn’t a good sign, but at least for now the economy appears robust.


The argument that the AI bubble has already burst is plausible, but only if this was a mini-bubble, or collection of them around data centers and the chips used in them.


Certainly, behavior usually associated with extreme froth was obvious in memory-chip stocks, with Korea a standout that sucked in leveraged retail buyers right at the top. Stock price ramps and crashes there look like the blowout final phase of most bubbles, but compressed into just a few months from the low in March after the U.S.-Israeli attacks on Iran to a high in June.


The good news is that a lot of this mini-bubble has deflated. Chip maker Micron almost quadrupled in three months, before falling by more than half. Semiconductor companies more broadly doubled, then fell 43%. Suppliers of energy and equipment to build AI data centers showed the same pattern, in the case of Bloom Energy with just as extreme moves.


The same goes for the weak “neocloud” data-center operators, the subprime borrowers of this cycle, notably Nvidia-backed CoreWeave. CoreWeave shares more than quadrupled after its IPO last year, but have since sunk all the way back.



Further falls to get back to the level at the start of the year would be painful, but nothing like the 2001 selloff, when router-maker Cisco, already down by half, halved again in the first three months—and didn’t recover its March 2000 high until December last year.


The bad news is that this might not be merely a mini-bubble. The broader AI trade is worryingly circular. Falling shares might lead companies to pull back their aggressive investment plans, hitting suppliers that have also been bid up, hurting their lenders and feeding back into still-lower demand. If this happens, expectations of extraordinary growth in data-center spending will be cut, hitting the economy as well as the companies that investors had hoped would profit from it.


Lower share prices could also spill over into other sectors. Retail investors have been big buyers of stocks in general and AI stocks in particular, and economists think their profits from rising stocks contributed to strong consumer spending. If that wealth effect goes into reverse, it will weaken the economy and hurt companies that have nothing to do with AI.


Vincent Mortier, chief investment officer of fund manager Amundi, says: “Generally you know you’re in a bubble when it bursts.” So far all we can say for sure is that a mini-bubble has popped in data centers and their suppliers, and that some of the speculative froth has been wiped from much of the rest of the AI trade.


Whether the air is coming out of a bigger AI bubble is impossible to know—but scarily plausible.

 
 
 

Recent Posts

See All

Comments


  • Facebook
  • Twitter
  • LinkedIn

©2021 by The Spritzler Report. Proudly created with Wix.com

bottom of page