Gregmal Posted June 6 Posted June 6 28 minutes ago, kab60 said: A lot of the earnings are circular though. Nvidia books earnings upfront, customers depreciate these things over 5-7-10 years. Of course you get a massive spike in earnings. Question is what follows. Sure dotcom had silly companies, but take a look at Tesla/SpaceX. The magnitude is beyond anything we have ever seen. Yea I think this is 1999 new tech enthusiasm, coupled with a covid earnings type phenomena occurring thats giving folks excuses to justify paying for it. I mean even at the highest levels, Bill Ackmans basically running a closet MAG7 fund and Dan Loeb just pitched NVDA as cheap.
Spekulatius Posted June 7 Posted June 7 10 hours ago, 73 Reds said: Yeah but dot com was worse. At least AI companies generate sales and earnings. Some of you may not have been around but literally anything not related to tech was left for dead and 401k "millionaires" all believed they were investment geniuses; who needed stocks that didn't go up 10%/week? OpenAI and Anthropic almost certainly don’t make money, but they have sales. The companies that did do the buildout like JDSU, Cisco, Ciena (note they are all back now) did make money and lots of it. Where do you see the difference? We have now SPCX, a bunch of quantum computing stocks (zero revenue ). Memory prices up 4-5x (worse than in the dot do boom when they also went up but for less). AI Capex as a percentage of GDP is 2-3x the % of Capex we had during the fiber optic boom. So I am not sure the dot com boom was worse. I think what we currently have is bigger.
tnathan Posted June 7 Posted June 7 On 5/20/2026 at 1:56 PM, Red Lion said: On this subject I just read an opinion piece somewhere (maybe WSJ or Bloomberg?) which was saying small caps and emerging markets may be the best way to play the AI wave as there is the most room for margin expansion, and these tools are being made available to everyone (not just Fortune 500 companies). I found it to be an interesting thesis, this article was suggesting indexes rather than actual companies, but I'm sure there may be lots of smaller less efficient companies that can really improve their productivity and margins with AI tools. AI will only increase margins for companies in industries where competitive intensity is low / where the company already has a moat. If a bunch of retailers all use AI to cut headcount needs, ultimately those savings will be passed to consumers if there is a lot of competition. Small caps in general tend to have less moats, so wouldn't be surprised if in the medium and longer term this didn't happen.
kab60 Posted June 7 Posted June 7 4 hours ago, tnathan said: AI will only increase margins for companies in industries where competitive intensity is low / where the company already has a moat. If a bunch of retailers all use AI to cut headcount needs, ultimately those savings will be passed to consumers if there is a lot of competition. Small caps in general tend to have less moats, so wouldn't be surprised if in the medium and longer term this didn't happen. That's my working assumption as well. It's not like department stores became more profitable when the escalator was invented. Everybody eventually adopted it, and consumers enjoyed the surplus. That seems to be the way it goes with new technology in most industries. Obviously with the caveat that execution matters, and some will be better and worse at adapting. I've been buying a lot of Autotrader in UK - an online marketplace for used car, which is basically a monopoly and a royalty on used car sales ( @changegonnacome that's one example of a good business at ~10x '26 profits assuming they keep up the buyback at these levels. They bought 0.6% of shares last week...). I didn't buy Autotrader because of AI, but I think they're an example of a business that should be able to keep some of the economic gains for itself. Both directly, on the cost side, but also in the form of launching more features/increased product velocity, which improves ability to take price. In the same vain, I think investors sometimes focus overly much on new technology vs. the power of distribution/sticky installed base. If AI increases product velocity meaningfully, it should help (well-run) companies with a large installed base of customers but perhaps not the best tech/product. Everything is tbd, but that could be someone like Global Payments.
John Hjorth Posted June 7 Posted June 7 (edited) 12 hours ago, SharperDingaan said: +1 The only reason we're punting on SpaceX is because we're sure we can rely on the greed. A great deal of money rests on this historic trillion dollar IPO not only doing well, but also closing the day at well above the issue price. A material bump that also bumps the indices, and opens the way for those lined up behind it .... and the fees that come with it. Heaven and earth moved to ensure that nothing threatens it. The GFC demonstrated that excessive fee income corrupted the judgement of rating agencies and associated analysts. We would expect that the same thing is happening here. SD Yeah, tell us about it, SD [ @SharperDingaan ], Just a tiny report from the European side of the Atlantic pond here - a screen shot of the entry page in the app for my primary investment bank : I certainly haven't signed up for getting spammed like this, in fact I find it appalling, and insulting in a way. It's all about attracting more money and new customers. I'll file a complaint by phone tomorrow about it. From what I have read, the sale is handled here in Europe by : Nordnet Bank, Saxo Bank, Interactive Brokers & Revolut. Edited June 7 by John Hjorth Fixed some typos
SharperDingaan Posted June 7 Posted June 7 1 hour ago, frommi said: These banks must get big fees, i also get spammed a lot. Assume 1% of the total raise .... an awful lot of bribe money spread about, and the largest commission of the entire year. But ... it has to be successfull. SD
Spekulatius Posted June 7 Posted June 7 47 minutes ago, SharperDingaan said: Assume 1% of the total raise .... an awful lot of bribe money spread about, and the largest commission of the entire year. But ... it has to be successfull. SD What does success mean though? If it gets issues at $135 and all the shares are sold, then it stays to $135 the same day and keeps dropping to $80 in the coming month, the offering was successful for Elon, the underwriting banks. It just wasn’t a good investment for the participants. This happened with the Facebook IPO by the way and that was way cheaper but similarly hyped.
SharperDingaan Posted June 7 Posted June 7 (edited) 1 hour ago, Spekulatius said: What does success mean though? If it gets issues at $135 and all the shares are sold, then it stays to $135 the same day and keeps dropping to $80 in the coming month, the offering was successful for Elon, the underwriting banks. It just wasn’t a good investment for the participants. This happened with the Facebook IPO by the way and that was way cheaper but similarly hyped. There is a need to hype the market for those IPOs that follow, and ensure that price doesn't fall below the underwriters guarantee for the two week period following launch day. Intentionally underprice the IPO so that it pops 30%+ on issue day. Apply the marketing engine to drive price up 100% within the two weeks following. Creation of an options market to sell puts to investors, and calls to speculators. Not their problem if investors don't address their elevated risk, whether it is via the stock itself, the indices that are forced to buy it, or the options market. Of course, if a Spacex launcher now explodes on the pad .... it is 'kaboom' in more ways than one. So .... no launches for the next few weeks ... and 'controlled video'. There are ways around the indices problem. Back in the day we had Nortel, and later RIM, doing the same thing. Shitty way of doing business, but one can either vent, or use the opportunity. Just keep the greed in check, and ensure that you engage a mechanism by which to keep the funny money. Most will not, and piss it away .... believing they are invincible. Another opportunity to harvest. SD Edited June 7 by SharperDingaan
John Hjorth Posted June 7 Posted June 7 (edited) 3 hours ago, frommi said: These banks must get big fees, i also get spammed a lot. 2 hours ago, SharperDingaan said: Assume 1% of the total raise .... an awful lot of bribe money spread about, and the largest commission of the entire year. But ... it has to be successfull. SD I simply had to dig in, to find some information about it! [Getting appalled now and then keeps the blood pressure up! ] Prospectus, p. 47 & 48 : That's not that bad, is it? Then I started studying the specification further. USD 700 K in printing costs related to bringing mankind out in the galaxy gave me a good laugh! [<-John, the nitpicker!] Impressive part of that sum to helpers [accountants and attorneys] : USD 37,700 K! Then I stumbled and focused on the USD 4,000 K figure related to the European retail offering, combined with that this post releates only legal fees and expenses! -So, where are all the bank fees in specifification? Then I started reading paragraph 3.2. from the beginning, from top to end, and realized to get information about bank fees and commisions, I had to read paragraph 8.3.2 , ref. : Quote ... For more information on discounts, commissions and expenses, please refer to “8.3.2 Underwriting Discounts and Commissions”. ... Beginning of paragraph 8.3.2 : Quote 8.3.2 Underwriting Discounts and Commissions The Company and the underwriters have not yet agreed on the underwriting discounts and commissions. The underwriting discounts and commissions will be determined by the Company at pricing and will be agreed with the underwriters in the underwriting agreement. The Company expects that the total underwriting discounts and commissions for the global offering will be approximately $500 million. ... - - - o 0 o - - - When Bernard Arnault in August presents half year results for LVMH, I expect him talking about the experience of a sudden temporary spike in the sales of champagne just after the IPO of SpaceX! - - - o 0 o - - - Warner Bros. Entertainment [January 22nd 2013] : Cabaret | "Money" Musical Number | Warner Bros. Entertainment Edited June 7 by John Hjorth
SharperDingaan Posted June 7 Posted June 7 Great choice of video! For all those 'helpers' ... Master of the house, Les Miserables SD
gfp Posted June 7 Posted June 7 Like a broken record - I still recommend a Sunday listen to Jordi Visser's free weekly videos https://www.youtube.com/@JordiVisserLabs/videos
73 Reds Posted June 7 Posted June 7 15 hours ago, Spekulatius said: OpenAI and Anthropic almost certainly don’t make money, but they have sales. The companies that did do the buildout like JDSU, Cisco, Ciena (note they are all back now) did make money and lots of it. Where do you see the difference? We have now SPCX, a bunch of quantum computing stocks (zero revenue ). Memory prices up 4-5x (worse than in the dot do boom when they also went up but for less). AI Capex as a percentage of GDP is 2-3x the % of Capex we had during the fiber optic boom. So I am not sure the dot com boom was worse. I think what we currently have is bigger. There were companies without a single sale that attached ".com" to their corporate names and traded at ridiculous valuations. In fact there was really no way to value them at all yet folks were buying hand over fist. Most of those companies don't even exist today. I remember CSCO very well; you could sell a relatively short dated covered call option for the whole share price at one time. If that's not speculation, what is? I sort of get Space X; you're buying imagination as to what can possibly be but is yet unknown. Personally, I like a little of that in most of my equity investments.
tnathan Posted June 7 Posted June 7 10 hours ago, kab60 said: That's my working assumption as well. It's not like department stores became more profitable when the escalator was invented. Everybody eventually adopted it, and consumers enjoyed the surplus. That seems to be the way it goes with new technology in most industries. Obviously with the caveat that execution matters, and some will be better and worse at adapting. I've been buying a lot of Autotrader in UK - an online marketplace for used car, which is basically a monopoly and a royalty on used car sales ( @changegonnacome that's one example of a good business at ~10x '26 profits assuming they keep up the buyback at these levels. They bought 0.6% of shares last week...). I didn't buy Autotrader because of AI, but I think they're an example of a business that should be able to keep some of the economic gains for itself. Both directly, on the cost side, but also in the form of launching more features/increased product velocity, which improves ability to take price. In the same vain, I think investors sometimes focus overly much on new technology vs. the power of distribution/sticky installed base. If AI increases product velocity meaningfully, it should help (well-run) companies with a large installed base of customers but perhaps not the best tech/product. Everything is tbd, but that could be someone like Global Payments. GPN is interesting. I think GPN / FIS / FISV are cases of AI could hurt OR help them but probably too early to know. Either they can increase their own product velocity and re-write old tech stacks internally to keep current customers or customers can more easily rip and replace with modern software because AI makes that process more simple.
Spekulatius Posted June 7 Posted June 7 34 minutes ago, tnathan said: GPN is interesting. I think GPN / FIS / FISV are cases of AI could hurt OR help them but probably too early to know. Either they can increase their own product velocity and re-write old tech stacks internally to keep current customers or customers can more easily rip and replace with modern software because AI makes that process more simple. Re- writing existing tech stacks with AI is extremely hard, I think. It’s way easier to add features to existing software using AI. I believe AI will widen the gap between the tech have and the tech have nots. FIS could be one that I see adding new features to their “core” banking software . I bought some shares as @kab60 made some good arguments versus FISV. It’s still early on that one though.
Spekulatius Posted June 7 Posted June 7 1 hour ago, 73 Reds said: There were companies without a single sale that attached ".com" to their corporate names and traded at ridiculous valuations. In fact there was really no way to value them at all yet folks were buying hand over fist. Most of those companies don't even exist today. I remember CSCO very well; you could sell a relatively short dated covered call option for the whole share price at one time. If that's not speculation, what is? I sort of get Space X; you're buying imagination as to what can possibly be but is yet unknown. Personally, I like a little of that in most of my equity investments. So,what. the same thing happens now with AI. We have a shoe company (All odds) that pivoted into AI. Same with some crypto business with little or no revenues. We have exactly the same thing going on now but at a larger scale. Meme stocks. Money losing trillion $ IPO were not on the menu in 1999/2000 but there are several now. https://www.cnbc.com/2026/04/15/allbirds-bird-stock-shoes-ai.html
kab60 Posted June 7 Posted June 7 3 hours ago, Spekulatius said: Re- writing existing tech stacks with AI is extremely hard, I think. It’s way easier to add features to existing software using AI. I believe AI will widen the gap between the tech have and the tech have nots. FIS could be one that I see adding new features to their “core” banking software . I bought some shares as @kab60 made some good arguments versus FISV. It’s still early on that one though. Yeah, I am not highly confident in how it plays out, so it is not central to my thesis. In general, I do think investors often overestimate product/tech and underestimate distribution. Distribution and selling just ain't sexy.
73 Reds Posted June 7 Posted June 7 3 hours ago, Spekulatius said: So,what. the same thing happens now with AI. We have a shoe company (All odds) that pivoted into AI. Same with some crypto business with little or no revenues. We have exactly the same thing going on now but at a larger scale. Meme stocks. Money losing trillion $ IPO were not on the menu in 1999/2000 but there are several now. https://www.cnbc.com/2026/04/15/allbirds-bird-stock-shoes-ai.html 3 hours ago, Spekulatius said: So,what. the same thing happens now with AI. We have a shoe company (All odds) that pivoted into AI. Same with some crypto business with little or no revenues. We have exactly the same thing going on now but at a larger scale. Meme stocks. Money losing trillion $ IPO were not on the menu in 1999/2000 but there are several now. https://www.cnbc.com/2026/04/15/allbirds-bird-stock-shoes-ai.html Meme stocks and crypto long preceded the current AI craze. I agree with you about both of those; largely without value in their present iterations. Yet if companies in existence today can become more profitable through AI, valuations should rise. Not saying they all make sense but "dot com" was an entirely different animal.
Sionnach Posted June 7 Posted June 7 (edited) Great article on how the AI Boom is more similar to 2008 than the Dot com bubble that everyone is comparing it to https://groundbreakerre.substack.com/p/peak-cheap-the-ai-boom-isnt-2000 Quote The argument of this article is that the AI and semiconductor complex is an earnings bubble of exactly this kind - that its earnings are not merely cyclical but partly circular and partly an accounting choice - and that, like 2008 and unlike 2000, the whole structure is being financed with debt against collateral that does not behave the way the debt assumes. Get the diagnosis wrong and you will buy the cheap multiple and call it value. It’s not value. It’s peak cheap. Edited June 7 by Sionnach
Gregmal Posted June 8 Posted June 8 4 hours ago, Sionnach said: Great article on how the AI Boom is more similar to 2008 than the Dot com bubble that everyone is comparing it to https://groundbreakerre.substack.com/p/peak-cheap-the-ai-boom-isnt-2000 Yea I think theres bits of pieces of the last few manias here. Dot com new tech excitement, 08 accounting, 2021 supply/demand distortion/over earning.
Libs Posted June 8 Posted June 8 11 hours ago, gfp said: Like a broken record - I still recommend a Sunday listen to Jordi Visser's free weekly videos https://www.youtube.com/@JordiVisserLabs/videos gfp my friend...... It's possible I hate money, because Visser has made some great calls and I've missed every one of them. (BTW he also had a solid insight in last week's clip about using 200-day averages as a buying / selling signal. I'm going to look into that. He basically showed how much of the market collapses you could have avoided simply by using that tool and acting accordingly.) I have a more narrow question for you (or anyone) that pertains to the AI bubble (if it is one). It has been said that AI bears simply don't use LLM models, and thus 'don't get it.' I know you are fully engaged in using LLM's and vouch for them. Ditto Visser. But here's the rub. When I ask enthusiasts of Claude / Gemini Open AI etc. how much they are paying, it's always "zero" or "$20 a month." And when I ask how much more they would pay, I get a blank stare or 'nothing.' So how does this ecosystem continue when the product is being sold at a fraction of the cost? Here's a tl/dr: https://ea.rna.nl/2026/06/07/anthropic-openai-may-be-spending-more-than-1000-for-every-100-you-pay-them/?utm_source=substack&utm_medium=email And how much would you be willing to pay for Claude or whatever you use (this is an open question to all). I still say this all ends in tears, in terms of AI - related stocks. Just like every technological / industrial revolution (from an investing standpoint).
Gamecock-YT Posted June 8 Posted June 8 4 hours ago, Libs said: (BTW he also had a solid insight in last week's clip about using 200-day averages as a buying / selling signal. I'm going to look into that. He basically showed how much of the market collapses you could have avoided simply by using that tool and acting accordingly.) You'd think if he called every bubble in the past 40 years he wouldn't be having to do youtube videos and shill on the investor advisor circuit. "Oh, look it's easy! Just sell when the red line crosses the green line!" Totally ignoring that the signal had came up 2-3 times prior to that and kept going up.
Paarslaars Posted June 8 Posted June 8 (edited) 5 hours ago, Libs said: And how much would you be willing to pay for Claude or whatever you use (this is an open question to all). Isn't this similar to Facebook at the time? How will it make money, is someone ever going to pay to use facebook? And then it turned out they can make massive revenue via ads... They will find a way to properly monetize on AI, if not via subscriptions it will be via data mining & selling that data or other ways. 34 minutes ago, Gamecock-YT said: You'd think if he called every bubble in the past 40 years he wouldn't be having to do youtube videos and shill on the investor advisor circuit. "Oh, look it's easy! Just sell when the red line crosses the green line!" Totally ignoring that the signal had came up 2-3 times prior to that and kept going up. While I agree on the charts thing, I don't think he 'has to' do these videos. He's done well enough... I also don't think we are in a rotation yet, the whole market went down on Friday, not just AI but also oil, silver, gold, btc, etc... this was just a liquidity drain due to rate hike fears. Wouldn't surprise me if we end up green today. Edited June 8 by Paarslaars
thowed Posted June 8 Posted June 8 I'm happy paying for Claude atm as it is operating as my 'intern' i.e. writing intital reports on companies I'm researching. I would like to spend more time understanding what I can get out of it, as I am scratching the surface I think. I have no idea what will happen with AI, like I had no idea what would happen with the internet in 1997. What often happens with massive infrastructure spends is that they're overrated in the short-term and underrated in the long-term. So I expect the share prices to go down as they're overly optimistic at the moment. I also suspect (but total speculation) that AI will become, like the mobile phone companies in the late 90s, a Utility and will command valuations like that, rather than tech. (I don't remember the US situation, but in Europe Vodafone was the biggest stock in town in the late 90s).
John Hjorth Posted June 8 Posted June 8 (edited) On 6/6/2026 at 10:56 PM, changegonnacome said: Nasdaq prostituted itself via index inclusion rule changes to secure the listing over NYSE…..there is something disheartening around an institution like Nasdaq doing something so brazen, low down, obvious and clearly hostile to the end investor which they claim to serve to make a few bucks and score some points. Stuff like this was always done in the shadows, the skim counted in millions, not billions…see the participants were embarrassed to be perceived as being so greedy, so low down that they hid it well….it’s says a lot about the current moment that the participants of this scheme were happy to do it in broad daylight (indeed everybody is talking about it it’s so obvious). @changegonnacome, That you for the elaboration. Edited June 8 by John Hjorth
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now