OT: Nvidia/AI/Tech Exposure

horshack.sixpack

All-American
Oct 30, 2012
11,688
8,721
113
I ran some analysis on my portfolio which led me to sell my individual holdings of NVDIA and INTC today. Basically, I found that nearly every fund that I own has substantial exposure (basically 40%+ of my overall portfolio). AI is going to pop at some point. I'm not a believer in market timing, but I do believe in diversification. I was surprised how exposed I was/and still am to big tech. Also worth noting that AMD, Google, Amazon, META and Microsoft are all heavily invested in their own chips to, at a minimum, reduce their dependence on NVIDIA. Toss in Broadcom, TSMC (OEM chip manufacturer for many) and Marvell and I could see NVIDIA dominance drop from 80-90% to 50-60%. Still a great business but big stock hit. None of these companies will be content to cede the level of market control that NVIDIA has, indefinitely.

HoldingExposure to NVDA/MSFT/AAPL/META/AMZN
VOOHigh
QQQMVery High
VUGVery High
VONGHigh
NVDADirect single-stock exposure
 

JackReacherDawg

Sophomore
Apr 7, 2026
213
120
43
I ran some analysis on my portfolio which led me to sell my individual holdings of NVDIA and INTC today. Basically, I found that nearly every fund that I own has substantial exposure (basically 40%+ of my overall portfolio). AI is going to pop at some point. I'm not a believer in market timing, but I do believe in diversification. I was surprised how exposed I was/and still am to big tech. Also worth noting that AMD, Google, Amazon, META and Microsoft are all heavily invested in their own chips to, at a minimum, reduce their dependence on NVIDIA. Toss in Broadcom, TSMC (OEM chip manufacturer for many) and Marvell and I could see NVIDIA dominance drop from 80-90% to 50-60%. Still a great business but big stock hit. None of these companies will be content to cede the level of market control that NVIDIA has, indefinitely.

HoldingExposure to NVDA/MSFT/AAPL/META/AMZN
VOOHigh
QQQMVery High
VUGVery High
VONGHigh
NVDADirect single-stock exposure
The AI bubble will pop any day now. Every day it becomes more apparent that the profit potential is a fraction of current valuations. It's just a question of when the herd makes its collective move.

Nvidia chips will be hit the hardest. When the smoke clears and one or two LLM companies are left standing, they wont have enough buyers for the current supply.
 

horshack.sixpack

All-American
Oct 30, 2012
11,688
8,721
113
The AI bubble will pop any day now. Every day it becomes more apparent that the profit potential is a fraction of current valuations. It's just a question of when the herd makes its collective move.

Nvidia chips will be hit the hardest. When the smoke clears and one or two LLM companies are left standing, they wont have enough buyers for the current supply.
Well, considering regular companies can't get basic servers or memory right now because all these data center customers are content to pay list, thus sucking up all production from companies that have no incentive to sell it to their long term customers, the secondary market may find some takers at a fraction of the cost when the winners and loser sort it out. The tech market is a disaster right now as far as lead times. I had baked in better supply chain risk controls as a part of lessons learned from COVID, but even that is not enough to compensate for the lack of available gear rn.
 
  • Like
Reactions: patdog

horshack.sixpack

All-American
Oct 30, 2012
11,688
8,721
113
Did something similar with my mom's portfolio today. Van Eck SemiConducter had gone up so much it was a huge part of her total portfolio. So we sold about 20% of her total portfolio and moved it into a dividend stock fund.
I ran a heavy dividend portfolio for quite a while. I still have many of those stocks and they've actually been very good for growth and dividends. I'm currently bolstering my cash position to ideally have a year in reserve (liquid), then I'll restart my pursuit of new dividends.
 
  • Like
Reactions: patdog
May 20, 2023
210
324
63
I’m far from an investment expert but a lot of the bulls feel that this AI era is so different from anything we’ve experienced that investors A. Don’t understand it and B. Are scarred from the dotcom bubble so they don’t believe there is anyway that these stocks can run so well. I bought in to a lot of the AI “picks and shovels” about a year ago and did pretty well up until a couple of weeks ago. I trimmed a lot of what I owned early last week so I didn’t get his as hard as those that held firm. What makes it all so confusing is that there are good arguments from both sides. Nvidia is really confusing; it has had crazy earnings numbers and for more than a year hasn’t moved up at all. I’m getting close to going back to etf’s and t’bills. Yeah, I’m novice.
 

patdog

Heisman
May 28, 2007
59,156
29,593
113
I’m far from an investment expert but a lot of the bulls feel that this AI era is so different from anything we’ve experienced that investors A. Don’t understand it and B. Are scarred from the dotcom bubble so they don’t believe there is anyway that these stocks can run so well. I bought in to a lot of the AI “picks and shovels” about a year ago and did pretty well up until a couple of weeks ago. I trimmed a lot of what I owned early last week so I didn’t get his as hard as those that held firm. What makes it all so confusing is that there are good arguments from both sides. Nvidia is really confusing; it has had crazy earnings numbers and for more than a year hasn’t moved up at all. I’m getting close to going back to etf’s and t’bills. Yeah, I’m novice.
I’m no expert. But I don’t see long term
Demand for chips do anything but go up. Way up. But it had gotten to the point 50% of mom’s portfolio was in Van Eck Semiconductor. And another 25% in QQQ , which is also fairly heavy in the same sticks. So moving about 25% into something more stable seemed like a good idea. Especially with not in my backyard movement in USA right now. I could see a real short term pullback in these tech stocks. And she still about 25% in Van Eck so she’s got plenty of potential if it does keep going up.
 

ETK99

Heisman
Jul 30, 2019
10,020
13,774
112
What AI are you using, and what version? I'm fortunate to have access to some high level AI versions now and I ran some things as well. Everyone should be seeing solid market returns right now, but it got me 4% more than I was making.
AI is moving at a crazy rate, like Goat level crazy!
 
Last edited:
  • Like
Reactions: Papapat.sixpack

MissippiMade

All-Conference
Sep 13, 2024
604
1,498
93
What ETF’s are yall diversifying into? So hard to find solid ETF’s that don’t have tech as the main holdings unless you pick industry specific ETF’s etc, that’s what I’ve been struggling with
 

mstateglfr

All-American
Feb 24, 2008
16,521
6,336
113
Well, considering regular companies can't get basic servers or memory right now because all these data center customers are content to pay list, thus sucking up all production from companies that have no incentive to sell it to their long term customers, the secondary market may find some takers at a fraction of the cost when the winners and loser sort it out. The tech market is a disaster right now as far as lead times. I had baked in better supply chain risk controls as a part of lessons learned from COVID, but even that is not enough to compensate for the lack of available gear rn.

I have had to switch a couple orders a few months ago away from HP Order Built because HP just 17ing sat on the orders for a couple months and they said 'we can't fulfill these right now and there will be a price increase in July 1'.
I get that my orders for 30 high power computer lab bundles is drop in the bucket, but hearing they refused to even fulfill the orders was crazy.
And that's after I got over the price shock compared to a lab order we made in January.

It's just bonkers.
Apple jacking up the price of a base ipad by $100 was crazy. They held off for so long that it was basically 2 price increases at once.
A 30% price increase is steep(that's educational pricing) for schools.
 
  • Like
Reactions: jethreauxdawg

horshack.sixpack

All-American
Oct 30, 2012
11,688
8,721
113
I have had to switch a couple orders a few months ago away from HP Order Built because HP just 17ing sat on the orders for a couple months and they said 'we can't fulfill these right now and there will be a price increase in July 1'.
I get that my orders for 30 high power computer lab bundles is drop in the bucket, but hearing they refused to even fulfill the orders was crazy.
And that's after I got over the price shock compared to a lab order we made in January.

It's just bonkers.
Apple jacking up the price of a base ipad by $100 was crazy. They held off for so long that it was basically 2 price increases at once.
A 30% price increase is steep(that's educational pricing) for schools.
Dell either won’t quote servers that we need, or they quote and then cancel the order as soon as you place it. I have basic Meraki switches that have lead times substantially longer than Covid times.
 

57stratdawg

Heisman
Dec 1, 2004
148,513
24,294
113
Their product is more valuable than money.

I could see the case to diversify out of 1 Chip manufacturer (primarily $NVDA), but I’m not leaving the industry.
 
  • Like
Reactions: jethreauxdawg

JackReacherDawg

Sophomore
Apr 7, 2026
213
120
43
Their product is more valuable than money.

I could see the case to diversify out of 1 Chip manufacturer (primarily $NVDA), but I’m not leaving the industry.
My view is, if an industry is redonk overvalued, then that industry is redonk overvalued, so it's a losing play. Sure, one or two of these are going to hit. But I dont have a great idea which ones, so id have to invest in all or take a gamble. And if I invest in all then im guaranteed to not see my return.

 

Bilbo69

Junior
Nov 9, 2025
187
234
43
If you are young the only move is to park in the S&P 500. If you’re on the older side I’m getting the sense a crash is coming.
 

jethreauxdawg

Heisman
Dec 20, 2010
11,288
15,454
113
I’m gonna keep pouring (ok, it’s more of a drip) my money into fossil fuels. I understand that technology for the most part. Flammable stuff burns.
 

GloryDawg

Heisman
Mar 3, 2005
20,147
18,255
113
Until either the United States or China say uncle on AI they are going to keep pouring money into. As long as the money flows it keep on going. Just my humble opinion.
 
  • Like
Reactions: Papapat.sixpack

horshack.sixpack

All-American
Oct 30, 2012
11,688
8,721
113
I would never bet against Jensen Huang.
Nor would I. It will be interesting to see if the pie grows enough for NVIDIA to hold their valuation or if they become one of many viable alternatives. Regardless, all of these stock mutual funds own so much of that space that you will have substantial exposure to the sector without holding an individual position. That was my main point. I needed to not own it directly as well.
 

Perd Hapley

All-American
Sep 30, 2022
6,313
7,532
113
I think the question isn’t whether the boom can keep going (it probably can, for a few more years anyway), but rather who (or which specific sector) is left holding the bag at the end.

AI is like the biggest major infrastructure project the USA or the world has ever seen, with nothing close in 2nd place. But unlike every other major infrastructure endeavor, it’s almost entirely private sector driven. It’s touching everything….semiconductors, chips, electronics, automotive, energy, retail, arts / entertainment, and so forth. I’m no expert, but I think the further you go up through supply chain tiers, the more risk there will be.

Companies like MU and SK Hynix that have gigantic marketshare for HBM products seem well-positioned for the long term. They are selling to the hyperscaler cloud infrastructure / data center companies (sometimes directly and sometimes as pass-through via an NVIDIA, Oracle, etc.). Then you have your NVIDIA’s in the middle, who are selling to Microsoft and other Mag 7, but are having to purchase the memory capacity required by AI. Then finally you have the Mag 7 that are the interface between the end consumer and all of the above.

Once the “build out” is complete, eventually the Mag 7 will have a slowdown of data center demand. But, they’ll still have their existing profit centers to draw from (AWS, YouTube, Facebook, etc.), so they’ll be fine. The HBM providers will have less demand, but still some consistency with their direct sales to automotive and other continuous manufacturing goods. Where the squeeze might happen is the middle, where NVIDIA and others will have very little pricing power, less they lose existing contracts with the Mag 7. Anyone in the chip world that doesn’t have access to HBM production in-house carries a lot of risk at the current valuations, I think.
 

JackReacherDawg

Sophomore
Apr 7, 2026
213
120
43

Over $1.5T in debt they have off the books. Once the data centers that borrowing was for go live, they have to put the debt on the books. If the revenue projections aren't being met, then the party absolutely cannot continue (if it even gets to then).
 

JackReacherDawg

Sophomore
Apr 7, 2026
213
120
43
I think the question isn’t whether the boom can keep going (it probably can, for a few more years anyway), but rather who (or which specific sector) is left holding the bag at the end.

AI is like the biggest major infrastructure project the USA or the world has ever seen, with nothing close in 2nd place. But unlike every other major infrastructure endeavor, it’s almost entirely private sector driven. It’s touching everything….semiconductors, chips, electronics, automotive, energy, retail, arts / entertainment, and so forth. I’m no expert, but I think the further you go up through supply chain tiers, the more risk there will be.

Companies like MU and SK Hynix that have gigantic marketshare for HBM products seem well-positioned for the long term. They are selling to the hyperscaler cloud infrastructure / data center companies (sometimes directly and sometimes as pass-through via an NVIDIA, Oracle, etc.). Then you have your NVIDIA’s in the middle, who are selling to Microsoft and other Mag 7, but are having to purchase the memory capacity required by AI. Then finally you have the Mag 7 that are the interface between the end consumer and all of the above.

Once the “build out” is complete, eventually the Mag 7 will have a slowdown of data center demand. But, they’ll still have their existing profit centers to draw from (AWS, YouTube, Facebook, etc.), so they’ll be fine. The HBM providers will have less demand, but still some consistency with their direct sales to automotive and other continuous manufacturing goods. Where the squeeze might happen is the middle, where NVIDIA and others will have very little pricing power, less they lose existing contracts with the Mag 7. Anyone in the chip world that doesn’t have access to HBM production in-house carries a lot of risk at the current valuations, I think.
"They'll be fine". Yeah, but how far does the stock drop?

By traditional metrics the S&P500 is overvalued by 50% right now. Im not gonna buy the "this time is different" view while theres giant off-books debt and revenue is at 1% of projections.
 
  • Like
Reactions: patdog
Jul 5, 2020
546
475
63
Posted this in the wrong thread, but it's about defensive stocks, so-

I got defensive one year ago because I hated all of the market chop from AI and the admin juicing the market with the tariff BS. I positioned equally in UHC (bingo), PM and UPS for the dividends. I figured that since they are all decent recession plays, I'd be okay. Combined with VIG, it's performed steadily for me at about 18% YTD even though I could have made more with more tech exposure.

The UHC play was shrewd but I didn't expect it to do as well as it has in one year. Always a reminder to buy well-performing companies when there's a discount opportunity.
 

JackReacherDawg

Sophomore
Apr 7, 2026
213
120
43
Posted this in the wrong thread, but it's about defensive stocks, so-

I got defensive one year ago because I hated all of the market chop from AI and the admin juicing the market with the tariff BS. I positioned equally in UHC (bingo), PM and UPS for the dividends. I figured that since they are all decent recession plays, I'd be okay. Combined with VIG, it's performed steadily for me at about 18% YTD even though I could have made more with more tech exposure.

The UHC play was shrewd but I didn't expect it to do as well as it has in one year. Always a reminder to buy well-performing companies when there's a discount opportunity.
Close to same here. I moved heavy into dividend company funds a bit over a year ago, and it did well. Later I moved a little into utility companies, before the Iran War of Low IQ. That has not done well. At the same time I was looking to buy good companies at a discount (especially dividend payers), after kicking myself for not doing this with NVidia. I did well with Boeing, Ford, Target, Verizon, and Nucor and Steel Dynamics made a killing. I did well with a bit of gold and silver, but didnt sell so now im back where i started with it. I didn't do well with Pepsi and KHC. (I'm a big believer that KHC is severely undervalued, but am following Buffets advice to not buy in more than 3 times).

For a long while I was beating the market, even with a heavy bond allocation. Im not currently, though i'm close.

I have no idea what's a good buy right now. I've been building up cash :(. Will probably buy some more gold. Maybe 50/50 bonds and gold, one of the two has to do ok in a recession market right?
 

Perd Hapley

All-American
Sep 30, 2022
6,313
7,532
113

Over $1.5T in debt they have off the books. Once the data centers that borrowing was for go live, they have to put the debt on the books. If the revenue projections aren't being met, then the party absolutely cannot continue (if it even gets to then).
So that’s what….maybe an average of $350 billion each for GOOG, Meta, AMZN, and MSFT, and maybe $50~$100 billion for Oracle?

First off, here are the latest 2025 annual revenues for each of those companies: $717 billion, $403 billion, $318 billion, $201 billion, $67.3 billion. Expect all of those numbers to go up 20-35% before debt service begins on data centers. They have the capital to cover the debt without much trouble.

Secondly, it’s a bit naive to assume that the “off the books” debt isn’t already baked into the current stock price - like somehow that article you linked is some kind of secret info. Take Amazon for instance. CAGR of the stock for the past 5 years is a measly 5.8%. You’d damn near do better with treasury bonds. But they’ve grown revenues an average of over 13% over that period, and consistently beat their guidance for both revenues and EPS, too. So why isn’t the price following? Because they’ve continued to increase AI infrastructure spending.
 
  • Like
Reactions: JackReacherDawg
Jul 5, 2020
546
475
63
Close to same here. I moved heavy into dividend company funds a bit over a year ago, and it did well. Later I moved a little into utility companies, before the Iran War of Low IQ. That has not done well. At the same time I was looking to buy good companies at a discount (especially dividend payers), after kicking myself for not doing this with NVidia. I did well with Boeing, Ford, Target, Verizon, and Nucor and Steel Dynamics made a killing. I did well with a bit of gold and silver, but didnt sell so now im back where i started with it. I didn't do well with Pepsi and KHC. (I'm a big believer that KHC is severely undervalued, but am following Buffets advice to not buy in more than 3 times).

For a long while I was beating the market, even with a heavy bond allocation. Im not currently, though i'm close.

I have no idea what's a good buy right now. I've been building up cash :(. Will probably buy some more gold. Maybe 50/50 bonds and gold, one of the two has to do ok in a recession market right?
I'm with you on not knowing what to do next beyond holding serve.

RE: Precious Metals- I didn't go in for the gold buy although I certainly wanted to. I emotionally hedged with silver during its run-up, getting it at about $78. So, a bust right now but I didn't let it go and we'll see. I've been working very hard to understand the case for metals barring pure catastrophe. I'm either too dumb to understand it or too optimistic that the world's not ending.

The bond markets have also been unusual and I don't have much faith in that either.

I'm an accidental landlord from a house that I didn't sell when we upgraded several years back, and that's been my biggest refuge against middle-term inflationary fears, but I know that's not available for everybody. Also, being a landlord for one property is not ideal and the annual costs have soared.

I think that I've become comfortable with riding the market until it starts to clearly recede, at which time I'll happily take several years of gains against whatever loss it takes for me to see that *clear* recession point.

ETA that I am going to take a position in IBM because of the price shock which, in my experience, is overrated by 20-30%.
 

JackReacherDawg

Sophomore
Apr 7, 2026
213
120
43
So that’s what….maybe an average of $350 billion each for GOOG, Meta, AMZN, and MSFT, and maybe $50~$100 billion for Oracle?
No, they have about as much debt on the books. So double it.
First off, here are the latest 2025 annual revenues for each of those companies: $717 billion, $403 billion, $318 billion, $201 billion, $67.3 billion. Expect all of those numbers to go up 20-35% before debt service begins on data centers. They have the capital to cover the debt without much trouble.
I'm sure they can cover it. The question though is what happens to the stock price. And does the revenue justify the expenses.

I think they all know now that the current investment builds wont pay off (and most will never be built) but they're in a race to be the last one standing. Theyre pot committed. There's no off ramp.

I said before recently that I think the ones that can self fund this race will be the winners, even with inferior products. Alphabet, im looking at you.
Secondly, it’s a bit naive to assume that the “off the books” debt isn’t already baked into the current stock price - like somehow that article you linked is some kind of secret info. Take Amazon for instance. CAGR of the stock for the past 5 years is a measly 5.8%. You’d damn near do better with treasury bonds. But they’ve grown revenues an average of over 13% over that period, and consistently beat their guidance for both revenues and EPS, too. So why isn’t the price following? Because they’ve continued to increase AI infrastructure spending.
Fair point. But we must also remember that stock prices arent set by the average of what analysts think it should be. Its set by what the buyer thinks. If 100 people think a stock is worthless, and 1 thinks its worth $100....the stock price is $100, so long as he's buying. Sure, for large companies with tons of public equity this sort of settles out, the fan boys cant buy it all, but really we see that first with rising short bets, not price drops, right? And with all this self dealing and dumb passive money buying these company's stock....well maybe that makes the bubble. Did the fan boys drive it into top tier status, and then the dumb passive money keeps it there? Maybe.

The housing bubble data wasn't secret either. The market just had a different (exuberant) view. Until they didn't. My old adage i learned then still holds i think: a crisis takes longer to unfold then you'd think possible....and then it happens all at once. Held true with Covid too.
 
  • Like
Reactions: Perd Hapley

JackReacherDawg

Sophomore
Apr 7, 2026
213
120
43
No, they have about as much debt on the books. So double it.

I'm sure they can cover it. The question though is what happens to the stock price. And does the revenue justify the expenses.

I think they all know now that the current investment builds wont pay off (and most will never be built) but they're in a race to be the last one standing. Theyre pot committed. There's no off ramp.

I said before recently that I think the ones that can self fund this race will be the winners, even with inferior products. Alphabet, im looking at you.

Fair point. But we must also remember that stock prices arent set by the average of what analysts think it should be. Its set by what the buyer thinks. If 100 people think a stock is worthless, and 1 thinks its worth $100....the stock price is $100, so long as he's buying. Sure, for large companies with tons of public equity this sort of settles out, the fan boys cant buy it all, but really we see that first with rising short bets, not price drops, right? And with all this self dealing and dumb passive money buying these company's stock....well maybe that makes the bubble. Did the fan boys drive it into top tier status, and then the dumb passive money keeps it there? Maybe.

The housing bubble data wasn't secret either. The market just had a different (exuberant) view. Until they didn't. My old adage i learned then still holds i think: a crisis takes longer to unfold then you'd think possible....and then it happens all at once. Held true with Covid too.
Google tells me Alphabets 2025 proft was $130B on $400B in revenue. So if debt service becomes $50B a year, yeah that takes a big bite out of profits. And that debt service keeps going up.....

This is quickly getting beyond me, but I think that means revenues need to be hitting a trillion or so from AI alone in a couple years or so. Or else a major correction will happen.

Plus, how much of AI revenues comes at the expense of those current revenues? I dont think Google can sell ads via AI without it coming at the expense of ad revenue from Google search, etc. They already have ad revenue on steroids, so can they really see gains there? So thats a multiplier to needed revenues.
 

Perd Hapley

All-American
Sep 30, 2022
6,313
7,532
113
No, they have about as much debt on the books. So double it.

AMZN has a little over $200 billion in debt currently, against that $716 billion in annual revenue. Even if you give them $400 billion in new debt, they’ll be at 1.5x that total figure in annual revenue within 18 months. And they have more debt than anyone. Others - like GOOG as you mentioned, are in far better shape.

I'm sure they can cover it. The question though is what happens to the stock price. And does the revenue justify the expenses.

Going back to the example above, what’s the worst case for Amazon - the most highly leveraged company? Say they get to $900 billion annual revenue / $600 billion debt, and the profit turnaround from investment takes longer than expected. What happens? Stock levels out for 1-2 years at most? Stock drops 5% per year for 2 years? I think it’s more like the former. A 20-25% pullback…..yeah, that’s not happening.

Now, what’s the best case? Maybe they get to something more like $950 million revenue / $550 million debt, and turn profits quickly? What happens then? 15% bump every year moving forward? 20% per year? Both are possible. Upside seems to far outweigh risk, the way I see it.

I think they all know now that the current investment builds wont pay off (and most will never be built) but they're in a race to be the last one standing. Theyre pot committed. There's no off ramp.
Hmm….that’s quite a take. Not sure that I can really wrap my mind around the 5 or 6 largest and most lucrative companies on earth - each with the most brilliant and strategic minds on the planet - all simultaneously deciding to just have an expensive dick measuring contest with each other. Needless to say, I disagree.

I said before recently that I think the ones that can self fund this race will be the winners, even with inferior products. Alphabet, im looking at you.

I think it is near unanimous that they are ALL going to be winners….eventually. Who wins the most and who wins the soonest are the only question marks.

The elephant in the room - everyone saying “AI profits won’t justify the investments!” and repeating that broken record are not really aware of what’s coming. The reason why they aren’t aware of what’s coming is because the Mag 7 have been intentionally coy about that, because it’s certainly going to absolutely piss everyone off….while at the same time adding a costly necessity to everyone’s lives.

Where we are at now is the same place where we were in 1995 or 1996, when you could get a little computer program called Netscape, and get on this new fangled thing called “the internet”, for FREE, whenever you wanted to, for as long as you wanted. Then, capitalism conquered, and now it’s a $200+ required monthly expense for people to have high speed WiFi, plus a Smart Phone with high speed internet anywhere. And only 4 or 5 companies essentially control that entire market, in this country.

Now? ChatGPT and other similar free AI services are essentially Netscape. Some have already started trying to monetize it with monthly subscriptions. 10 years from now, one of the Mag 7 will be charging you out the asś for all of it. The free services will not be able to get the memory capacity or computing power required to keep up with these massive data-center driven entities backed by trillions in capital investment. You’ll be baited to pay Amazon $400 - $500 per month for an online shopping / phone / TV / internet bundle, with Alexa connecting it all and feeding you suggestions to each profit center. If you don’t like that, you can opt for the GFiber / YouTubeTV / Gemini package. Corporations are also going to pay out the asś for work suites that give them the bandwidth and the software needed to cut headcount and employees at an astonishing level.

Think about it like this - what if the internet just got invented, and you as a company could actually afford to build the infrastructure to control 25-30% of all the traffic? What would you do?

Fair point. But we must also remember that stock prices arent set by the average of what analysts think it should be. Its set by what the buyer thinks. If 100 people think a stock is worthless, and 1 thinks its worth $100....the stock price is $100, so long as he's buying. Sure, for large companies with tons of public equity this sort of settles out, the fan boys cant buy it all, but really we see that first with rising short bets, not price drops, right? And with all this self dealing and dumb passive money buying these company's stock....well maybe that makes the bubble. Did the fan boys drive it into top tier status, and then the dumb passive money keeps it there? Maybe.
80% of the market is institutional money. The price of the stock in this case is largely driven by what that 80% thinks. They compete on the margins, how big their beliefs are in Google vs. Amazon, Microsoft vs. Meta, etc. But you won’t find any of those 80% that think any of those above companies aren’t set up for the highest level of success in this new environment.

Don’t hear what I’m not saying though. I’m not calling for 50% stock price increase year over year for any of the Mag 7. I prefer the memory providers to the customer facing side for potential big returns. But they’ll all do reasonably well. This is just a generational type market shift where a new market is developing, and market share in that new environment simply costs what it costs.

The housing bubble data wasn't secret either. The market just had a different (exuberant) view. Until they didn't. My old adage i learned then still holds i think: a crisis takes longer to unfold then you'd think possible....and then it happens all at once. Held true with Covid too.

I wouldn’t describe my view as exuberant. I think its just the next level of what is required for these mega-cap companies to keep bringing in big returns, so they are simply doing what is necessary. The scale of work required to set up all these new superhighways opens up a lot of investment opportunity further down the supply chain, but in general they are just following the natural orders of what industry leaders do when new markets emerge.
 
Last edited:

JackReacherDawg

Sophomore
Apr 7, 2026
213
120
43
AMZN has a little over $200 billion in debt currently, against that $716 billion in annual revenue. Even if you give them $400 billion in new debt, they’ll be at 1.5x that total figure in annual revenue within 18 months. And they have more debt than anyone. Others - like GOOG as you mentioned, are in far better shape.
They turned $70B in profit. At $600B in debt, thats around half their profits....gone. And the debt service amount goes up every year, until AI revenue exceeds expenses.
Going back to the example above, what’s the worst case for Amazon - the most highly leveraged company? Say they get to $900 billion annual revenue / $600 billion debt, and the profit turnaround from investment takes longer than expected. What happens? Stock levels out for 1-2 years at most? Stock drops 5% per year for 2 years? I think it’s more like the former. A 20-25% pullback…..yeah, that’s not happening.
Thats nowhere near a worst case. The worst case is AI is unprofitable for them. Doesnt mean AI isn't a huge thing. Just means theres too much competition and expenses to turn a profit.

So, profits cut in half for decades. Thays not a 5% drop.
Now, what’s the best case? Maybe they get to something more like $950 million revenue / $550 million debt, and turn profits quickly? What happens then? 15% bump every year moving forward? 20% per year? Both are possible. Upside seems to far outweigh risk, the way I see it.
Sure, that's possible. The question is how does this best case map to every other AI player? How's Alphabet stock in this case? Tesla? Etc.
Hmm….that’s quite a take. Not sure that I can really wrap my mind around the 5 or 6 largest and most lucrative companies on earth - each with the most brilliant and strategic minds on the planet - all simultaneously deciding to just have an expensive dick measuring contest with each other. Needless to say, I disagree.
It happened with housing. Sort of.

Regardless, you are misstating my argument. My argument is they know, at best, only 1 or two players can survive this. At worst the competition and expenses make it unprofitable for anyone. But they all have incentives to keep in the race they got into already.
I think it is near unanimous that they are ALL going to be winners….eventually. Who wins the most and who wins the soonest are the only question marks.
Lol, no its not unanimous. Thats an insane statement. It doesnt math. At all.
The elephant in the room - everyone saying “AI profits won’t justify the investments!” and repeating that broken record are not really aware of what’s coming. The reason why they aren’t aware of what’s coming is because the Mag 7 have been intentionally coy about that, because it’s certainly going to absolutely piss everyone off….while at the same time adding a costly necessity to everyone’s lives.
Uh huh.
Where we are at now is the same place where we were in 1995 or 1996, when you could get a little computer program called Netscape, and get on this new fangled thing called “the internet”, for FREE, whenever you wanted to, for as long as you wanted. Then, capitalism conquered, and now it’s a $200+ required monthly expense for people to have high speed WiFi, plus a Smart Phone with high speed internet anywhere. And only 4 or 5 companies essentially control that entire market, in this country.
Or...3d TV. The meta verse. NFTs. Cable TV once streaming came around.
Now? ChatGPT and other similar free AI services are essentially Netscape. Some have already started trying to monetize it with monthly subscriptions. 10 years from now, one of the Mag 7 will be charging you out the asś for all of it. The free services will not be able to get the memory capacity or computing power required to keep up with these massive data-center driven entities backed by trillions in capital investment. You’ll be baited to pay Amazon $400 - $500 per month for an online shopping / phone / TV / internet bundle, with Alexa connecting it all and feeding you suggestions to each profit center. If you don’t like that, you can opt for the GFiber / YouTubeTV / Gemini package. Corporations are also going to pay out the asś for work suites that give them the bandwidth and the software needed to cut headcount and employees at an astonishing level.
Its certainly plausible. The problem is their product will require insane, ongoing costs, while "just-as-good" alternatives will be basically free. Businesses RIGHT NOW are abandoning OpenAI and Anthropic for cheap Chinese alternatives.
Think about it like this - what if the internet just got invented, and you as a company could actually afford to build the infrastructure to control 25-30% of all the traffic? What would you do?
I would look at the fundamentals, and see if they made sense. If they require simultaneous things that cant happen simultaneously....then no.
80% of the market is institutional money. The price of the stock in this case is largely driven by what that 80% thinks. They compete on the margins, how big their beliefs are in Google vs. Amazon, Microsoft vs. Meta, etc. But you won’t find any of those 80% that think any of those above companies aren’t set up for the highest level of success in this new environment.
Lol, its a common view and growing in this cohort that it's a bubble.
Don’t hear what I’m not saying though. I’m not calling for 50% stock price increase year over year for any of the Mag 7. I prefer the memory providers to the customer facing side for potential big returns. But they’ll all do reasonably well. This is just a generational type market shift where a new market is developing, and market share in that new environment simply costs what it costs.



I wouldn’t describe my view as exuberant. I think its just the next level of what is required for these mega-cap companies to keep bringing in big returns, so they are simply doing what is necessary. The scale of work required to set up all these new superhighways opens up a lot of investment opportunity further down the supply chain, but in general they are just following the natural orders of what industry leaders do when new markets emerge.
Ok
 

BoDawg.sixpack

All-Conference
Feb 5, 2010
5,593
3,162
113
Jamie Dimon was asked about AI recently.

----

"When I look at AI itself, the amount of money being spent is huge. Will it, in total, pay off? Probably, just like the internet did. Will it pay off the way you expect in the timetable you expect? Definitely not," Dimon said during a recent interview on the Master Investor Podcast with Wilfred Frost.

When asked if the market is pricing in a perfect outcome of the AI boom, Dimon said, "Not perfect, but probably a good outcome.

"It won't be without bumps and recalibrations," Goldman Sachs CEO David Solomon said of the AI boom during the bank's earnings call last week.

During the same podcast interview, Dimon described rocket maker SpaceX (SPCX) — one of the companies at the center of the spending boom — as "extraordinary."

He praised the company's satellite internet business and defended the feasibility of one of its more far-flung AI plays: space data centers.

"I've seen numbers on the data centers in space that can actually work … if it works, you're talking about very cheap energy, cheap cooling, very stable," Dimon said.
 
  • Like
Reactions: patdog

horshack.sixpack

All-American
Oct 30, 2012
11,688
8,721
113
Jamie Dimon was asked about AI recently.

----

"When I look at AI itself, the amount of money being spent is huge. Will it, in total, pay off? Probably, just like the internet did. Will it pay off the way you expect in the timetable you expect? Definitely not," Dimon said during a recent interview on the Master Investor Podcast with Wilfred Frost.

When asked if the market is pricing in a perfect outcome of the AI boom, Dimon said, "Not perfect, but probably a good outcome.

"It won't be without bumps and recalibrations," Goldman Sachs CEO David Solomon said of the AI boom during the bank's earnings call last week.

During the same podcast interview, Dimon described rocket maker SpaceX (SPCX) — one of the companies at the center of the spending boom — as "extraordinary."

He praised the company's satellite internet business and defended the feasibility of one of its more far-flung AI plays: space data centers.

"I've seen numbers on the data centers in space that can actually work … if it works, you're talking about very cheap energy, cheap cooling, very stable," Dimon said.
I'll be looking at SPCX in December when the 180 day period is up. Depending on where it is trading then, I might buy a little so I can set it next to my IBIT as a regrettable decision...
 
  • Haha
Reactions: jethreauxdawg

Perd Hapley

All-American
Sep 30, 2022
6,313
7,532
113
They turned $70B in profit. At $600B in debt, thats around half their profits....gone. And the debt service amount goes up every year, until AI revenue exceeds expenses.

IMG_1922.jpeg

They only turned $70B due to already investing over $50B of free cash flow into the AI / AWS infrastructure….same reason they took out debt (the amount of which is uncertain). You’re kind of double dinging them for the infrastructure investment. Once that debt is capped off and they go more into maintenance mode with the infrastructure, profit essentially goes from baseline $70B to baseline $120B, before accounting for any YoY growth from the current base business, or profitability from the AI investment. Account for all that, THEN you look at the debt service, and see what you get.

Back of the napkin math - you referenced a $70B profit being cut in half by the debt service, meaning a $35B annual debt cost on the books. Now, apply that same $35B to the $120B number above. Now you’re at $85B. With no growth at all YoY, you’re already looking at about a 20% profit increase just from restoring the FCF previously used for build out. If they keep generating the cash from the current core business over the short term, they’ll keep everything churning along. That simple.

Thats nowhere near a worst case. The worst case is AI is unprofitable for them. Doesnt mean AI isn't a huge thing. Just means theres too much competition and expenses to turn a profit.
The reason why they are all investing now is to remove the possibility of “too much competition”. As far as full scale AI driven business, really only AMZN and GOOG are going for broke. Meta is not really operating in the same space as those two, and likely will not require the same firepower. Anthropic will be an interesting litmus test for AI as a whole, but I question if even they can really compete at scale with AWS / Google, without any of the other supporting core business. They will have to get creative in a hurry, if they do in fact go public.

So, profits cut in half for decades. Thays not a 5% drop.
I don’t see that in the realm of possibility. But, half of $500 billion is $250 billion. If AMZN hits $1T revenue and $250 billion in profit in a few years, with another $250 billion in debt, I seriously doubt any of their investors will be complaining. That’s around $23 EPS. Their most recent report had a trailing 12 month EPS of $8.36. Say it falls somewhere more in the middle of range of outcomes, they’ve still doubled their EPS, even with all the debt.

Sure, that's possible. The question is how does this best case map to every other AI player? How's Alphabet stock in this case? Tesla? Etc.
I think they can all coexist, they don’t all do the same thing. Tesla - they may slip, their stock price has been outrageous compared to their fundamentals forever. But again, they aren’t a big player in this data center mega-build, anyway. There’s room for them in the space.

It happened with housing. Sort of.

Regardless, you are misstating my argument. My argument is they know, at best, only 1 or two players can survive this. At worst the competition and expenses make it unprofitable for anyone. But they all have incentives to keep in the race they got into already.
Define “survive”….are you seriously saying out of the 5-6 hyperscaler companies, that 3 or 4 are going to be literally gone in a few years due to not being able to make it work?

Lol, no its not unanimous. Thats an insane statement. It doesnt math. At all.

Uh huh.

Or...3d TV. The meta verse. NFTs. Cable TV once streaming came around.
I’m pretty sure I never went to work and had my Sr MGR tell my entire department at one of the largest companies in the world that we needed to start all becoming more proficient at using a 3D TV remote or a VR headset. As far as the cable / streaming example, they both suck.

You’re playing the doom and gloom / future obsolescence card on something that hasn’t even really taken off at all yet. While technically not wrong or right, you could make the same bear argument against any stock or entire asset class. It’s not a compelling discussion point. Will Costco still be as profitable 10 years from now, or will added competition from BJ’s and Sam’s Club curb their margins? Who the hell knows. I can make a completely baseless argument that they won’t be, and use the uncertainty of an absurd time horizon to back it up if nothing else.

Its certainly plausible. The problem is their product will require insane, ongoing costs, while "just-as-good" alternatives will be basically free. Businesses RIGHT NOW are abandoning OpenAI and Anthropic for cheap Chinese alternatives.
OpenAI and Anthropic need to be bigger than just a chatbot / code generator / algo generator company to make it. There’s no doubt about that.

But, AMZN and GOOG can actually buy the datacenter bandwidth to shut out any cheap alternatives. We’re largely talking about privitization of large swaths of internet bandwidth here. They’ll be able to block out competitors from their servers unilaterally. Its unheard of, but after net neutrality got the ax, it will be completely legal.

I would look at the fundamentals, and see if they made sense. If they require simultaneous things that cant happen simultaneously....then no.
It doesn’t require as many simultaneous things as you are stating. I don’t think we’re so far off from each other on this one. 1 or 2 megaplayers will emerge, then more niche players that need more specific use cases will also have room. There’s room for all of them in some capacity.
 

GloryDawg

Heisman
Mar 3, 2005
20,147
18,255
113
I would like to have the money China is investing in the misinformation campaign in the United States to get an edge.