The Investment Thread

[LouisEly]

 

 

[LouisEly]

On 9/6/2025 at 7:21 PM, BrewerFan said:

I think October is actually... one of the better months. August is one of the worst and Sept is the worst.

September is the worst.  But October is 4th worst and August is in the middle of the pack.

[BrewerFan]

45 minutes ago, LouisEly said:

September is the worst.  But October is 4th worst and August is in the middle of the pack.

The S&P500 didn't exist until the 1950s(I believe '57). 

So going back prior to that is estimating how it would have performed with the 500 companies that would have been in it... which isn't particularly useful for looking at a trend.

If you do start in '28 though, you're including the '29 crash. You still have the '87 crash and '08 crashes... though I don't think that really informs how you should invest, using 3 extreme outliers, but, it's not my money. 

October has historically been a very strong month, particularly if you're going to start a year before the worst crash in the last 100 years(and then two more outliers in '87 and '08) leading into the Christmas season. Unless you think a historic crash is coming, I think October is a pretty good month and... not one in which you need to trim 25%. 

 

TSM is up almost 28% since I posted that and Google about 8%. 

This is a better run down for long term investors. I do know trying to time the market and... beyond that, trying to use historic events to help inform that decision is... a poor idea. 

https://www.nasdaq.com/articles/should-you-worry-about-october-effect-heres-what-history-says

 

[BrewerFan]

On 8/6/2025 at 8:25 AM, BrewerFan said:

Anyway, both are getting beat up, but AMD had a lot of write offs as they haven't gotten their export license for China, BUT... they are still loading up on the MI300 for China(I think it's MI300). 

MI355 ramping and they're raising prices in Q3 due to demand and they have the MI400 coming out which compares favorably to the B200 for inference...kinda. 

 

On 9/6/2025 at 8:21 PM, BrewerFan said:

I am selling AMD, 

But the original premise was right. It won't take much to move AMD. Conversely... it's not going to take much to move NVDA DOWN but a LOT to move it up.

I love that everyone is sooo confident in AGI and that it could generate TRILLIONS over the next decade for NVDA, but I'm... not loving the circular investing.

OpenAI invests in AMD, NVDA invests in OpenAI... 1.5B.

Meanwhile the Govt' buys a steak in INTC and kinda...."encourages" NVDA to work with them so they can get the export controls lifted on the B30, but now China doesn't WANT to buy from them(though their largest companies still plan to if that does in fact end up happening). 

So Nvidia is investing 1.5B in a company that buys from Nvidia...

I haven't asked many questions yet, but that's getting a bit concerning(from a layman's viewpoint). Though, I will say, when it's been time to unload or buy something, I do get a call, so I'm just curious now. 

 

@owbcyou seem to be more knowledgeable on the tokenization and the GPUs infrastructure. Is this just Nvidia investing in another promising company or are they kinda... ensuring OpenAI has enough revenue to goose their revenue which is taking a potentially 50 BILLION dollar hit this year from the loss of China(which used to be 25% of their revenue)...

 

I will say, my inaction and being leery of paying the income tax vs long term Capital Gains on AMD has been an incredibly lucky and undeserved development. I had every intention of selling, I just... wasn't sure where else to put it. It's also got me wishing I'd bought 10K shares instead of 5K(then again, I get just sick looking at where I bought D-wave and where it is now... but I didn't have conviction in that investment). 

[nate82]

5 hours ago, BrewerFan said:

Meanwhile the Govt' buys a steak in INTC and kinda...."encourages" NVDA to work with them so they can get the export controls lifted on the B30, but now China doesn't WANT to buy from them(though their largest companies still plan to if that does in fact end up happening). 

The Govt' buying a steak in Intel is a soft bailout for Intel.  

NVDA basically bought out its third tier competition for dedicated consumer graphics cards.  Intel was failing in the consumer graphics card industry.  There was a market for them but they basically gave up and took NVDA's money to basically bail them out.  

Steam while not 100% accurate but it has the largest number of users doesn't even have the Intel ARC graphic cards in the top 25 for users.  NVDA dominates the consumer graphic cards market and even the low end graphic card market.  Intel tried to capture the low end but wasn't ever able to capitalize on it though they did improve this year.  

Intel still has a strong CPU presence but AMD has been gaining ground on them for the past 5-10 years now.  Intel used to be NVDA for CPU's where it had +70% of the market share.  

The non Arc intel ones are non dedicated GPU's.  

 

[BrewerFan]

5 hours ago, nate82 said:

The Govt' buying a steak in Intel is a soft bailout for Intel.  

NVDA basically bought out its third tier competition for dedicated consumer graphics cards.  Intel was failing in the consumer graphics card industry.  There was a market for them but they basically gave up and took NVDA's money to basically bail them out.  

Yeah, I'm not alll that concerned about the INTC part of the equation, that was just another kinda abnormal action. 

The Govt' taking a stake in American Companies? 

 

But that's not what I'm concerned with. I'm not even really all that concerned, but it's more with NVDA investing in it's clients in the DC market... it's clients investing in AMD... that CAN be a sign of a bubble... or they just really believe in OpenAI...

 

[wallus]

What's your favorite International ETF? I have been increasing my percentage of International and want to have more than a couple of ETFs.

[BrewerFan]

On 10/13/2025 at 11:28 AM, wallus said:

What's your favorite International ETF? I have been increasing my percentage of International and want to have more than a couple of ETFs.

I don't know, I stay away from international, but I was told VXUS...

 

Take that for what you will... but I came here just to say HOLY ####... NVDA hits 209 premarket.

 

Over a FIVE TRILLION  dollar Market Cap... 
That's with the Fed meeting tomorrow and with the jobs report... 25 or even 50 basis points seems very likely(25 seems likely, 50 seems a bit unlikely)... 

If there is a cut....AND a deal with China for rare earth minerals and we allow them to by our last gen GPUs(so AMD MI300, the H100 for NVDA)...this thing is gonna hit a 6T market cap in mid '26. 

I don't expect things to go that smoothly, but damn! I I mean, I'm waiting for a 10-15% "correction" or some recession in the next year or two, but AI may really be just unstoppable. AMD 265. AVGO 375, TSM 305. 

I'll assume there will be a lot of people taking profits after a run like this... just adding 500-600B in a few weeks. Not a bad thing to wake up to!

Also... not real since I will not be selling, but still, nice to see!

 

My youngest Nephew still wears diapers, but he's up 360%. Though I still don't have his SS# so... maybe his Uncle is really the one that's up!

[owbc]

6 hours ago, BrewerFan said:

I don't know, I stay away from international, but I was told VXUS...

 

Take that for what you will... but I came here just to say HOLY ####... NVDA hits 209 premarket.

 

Over a FIVE TRILLION  dollar Market Cap... 
That's with the Fed meeting tomorrow and with the jobs report... 25 or even 50 basis points seems very likely(25 seems likely, 50 seems a bit unlikely)... 

If there is a cut....AND a deal with China for rare earth minerals and we allow them to by our last gen GPUs(so AMD MI300, the H100 for NVDA)...this thing is gonna hit a 6T market cap in mid '26. 

I don't expect things to go that smoothly, but damn! I I mean, I'm waiting for a 10-15% "correction" or some recession in the next year or two, but AI may really be just unstoppable. AMD 265. AVGO 375, TSM 305. 

I'll assume there will be a lot of people taking profits after a run like this... just adding 500-600B in a few weeks. Not a bad thing to wake up to!

Also... not real since I will not be selling, but still, nice to see!

 

My youngest Nephew still wears diapers, but he's up 360%. Though I still don't have his SS# so... maybe his Uncle is really the one that's up!

The absurd CapEx spending on AI is propping up the economy right now. We have mass layoffs happening during an economic boom because they need to cut costs elsewhere else to pay for it. 

AI itself is not adding value (yet). The AI startups are bleeding money badly. Their infrastructure costs are through the roof. We have no ability to get electricity costs down in the short term. 

Obviously nobody can predict when corrections will happen, but in the short short tern it sure seems like things will continue to go up. 

[LouisEly]

24 minutes ago, owbc said:

The absurd CapEx spending on AI is propping up the economy right now. We have mass layoffs happening during an economic boom because they need to cut costs elsewhere else to pay for it. 

AI itself is not adding value (yet). The AI startups are bleeding money badly. Their infrastructure costs are through the roof. We have no ability to get electricity costs down in the short term. 

Obviously nobody can predict when corrections will happen, but in the short short tern it sure seems like things will continue to go up. 

I have access to a lot of consumer data..  Consumer spending has been going down.  That was before the federal shutdown.  Now you have hundreds of thousands of federal workers who are not getting paid who are going to drastically decrease spending.  While that's a fraction of the total employment base, it trickles across the economy eventually.  

As someone who has been laid off 3 times in the last 13 years, I can tell you that for every month you don't get a paycheck it takes at least two months to catch back up and feel comfortable returning to normal spending.  Unemployment maxes out at a little over $2K/month.  That won't even cover most mortgages.  

[owbc]

27 minutes ago, LouisEly said:

I have access to a lot of consumer data..  Consumer spending has been going down.  That was before the federal shutdown.  Now you have hundreds of thousands of federal workers who are not getting paid who are going to drastically decrease spending.  While that's a fraction of the total employment base, it trickles across the economy eventually.  

As someone who has been laid off 3 times in the last 13 years, I can tell you that for every month you don't get a paycheck it takes at least two months to catch back up and feel comfortable returning to normal spending.  Unemployment maxes out at a little over $2K/month.  That won't even cover most mortgages.  

It's a two-tiered economy right now. Higher earners are mostly doing fine. The bottom half to two-thirds is struggling to get by. 

At some point I suspect that half the country could be in poverty or borderline poverty and the economic metrics would continue to indicate a strong economy. 

[BrewerFan]

7 hours ago, owbc said:

The absurd CapEx spending on AI is propping up the economy right now. We have mass layoffs happening during an economic boom because they need to cut costs elsewhere else to pay for it. 

AI itself is not adding value (yet). The AI startups are bleeding money badly. Their infrastructure costs are through the roof. We have no ability to get electricity costs down in the short term. 

Obviously nobody can predict when corrections will happen, but in the short short tern it sure seems like things will continue to go up. 

No, but with each generation, the chips get more efficient. I think companies like Oklo will step in and provide power for some of these. 

And don't most start up's always bleed money? 
 

The mass layoffs... well, I'm skeptical they're needed to pay for AI. AMZN doesn't need mass layoffs to pay for AI, nor does MSFT or Norvo. Some companies are just struggling like Target, UPS... 

And Tariffs have also taken their toll. 

 

I'm not arguing AI is good for humanity. I think it's just the opposite... but it's the next industrial revolution, so as long as it's booming, I'm going to do all I can. Keep my money in NVDA, TSM, AVGO and AMD(though...I was VERY close to selling that not long ago). 

 

[homer]

I'd be curious to know how big an impact the advent of the internet had on economic numbers. AI might become a nothingburger when it's all said and done. Having said that, I'm glad I got in NVDA under $100. I absolutely think we're in some kind of bubble but impossible to say when it will burst.

[LouisEly]

24 minutes ago, homer said:

I'd be curious to know how big an impact the advent of the internet had on economic numbers. AI might become a nothingburger when it's all said and done. Having said that, I'm glad I got in NVDA under $100. I absolutely think we're in some kind of bubble but impossible to say when it will burst.

According to the research that I've read, AI so far amounts to about a 10% gain in efficiency.

That has wide variations depending on the specific application (coding for example, basic customer support, analytics, etc.) but the overall enterprise impact is about a 10% gain in efficiency.

[LouisEly]

Chipotle disappoints, down 15%, citing consumer weakness.

[owbc]

2 hours ago, LouisEly said:

According to the research that I've read, AI so far amounts to about a 10% gain in efficiency.

That has wide variations depending on the specific application (coding for example, basic customer support, analytics, etc.) but the overall enterprise impact is about a 10% gain in efficiency.

10% seems right based on my present use of Copilot for coding help. It’s definitely the best new tool since I switched to Python around 2014. 

Everyone is working as fast as possible on AI agents since that’s where the 50%+ efficiency gains could come. I have my doubts at their ability to replace software engineers. They seem more like customer service agent replacements. 

[nate82]

8 minutes ago, owbc said:

They seem more like customer service agent replacements. 

In the banking world it is being used to detect fraud.  Far easier for AI to sift through hundreds of thousands of transactions every day than it is for a human.  Even the applications that are running now struggle with this.

The fear of AI taking a lot of jobs is the same as the robotics fear.  Some jobs will be eliminated and some created.  Basic jobs will go away but more technical jobs and ones that need human input and interaction with won’t be as impacted.

Also LLM’s get mixed up with AI.  I use a LLM for nearly everything now mostly for emails.  I don’t have to spend hours a day reading or crafting replies to emails now.  The LLM can sift through it and give me a summary and I can put in a summary of what I want to respond and it will craft the email for me.  Though my LLM has a lot of India words like kindly that I have to get rid of or tell it not to use.

[igor67]

Is there an AI on the market that isn't built on training it's data using an LLM though?

[owbc]

18 minutes ago, igor67 said:

Is there an AI on the market that isn't built on training its data using an LLM though?

I’m referring to an LLM when I say AI.

All of what’s being built these days are LLM wrappers essentially. 
 

Using an LLM for emails is interesting. Email Is basically obsolete to me now, I doubt I write more than one email per day. 

[LouisEly]

1 hour ago, owbc said:

Everyone is working as fast as possible on AI agents since that’s where the 50%+ efficiency gains could come.

The 10% enterprise efficiency includes AI agents.  Like I said, specific applications will have far greater than 10% efficiency gains, but when all roles/responsibilities/tasks are taken into account, the overall net across the enterprise is 10% efficiency gain.

[nate82]

43 minutes ago, igor67 said:

Is there an AI on the market that isn't built on training it's data using an LLM though?

AI is a broad term like plants.  Are you talking about trees, shrubs or ?????

 

24 minutes ago, owbc said:

Using an LLM for emails is interesting. Email Is basically obsolete to me now, I doubt I write more than one email per day. 

I wish I only have to write one email per day.  But since I am in a highly regulated profession it is mostly emails all day everyday. 

[wallus]

My business is seasonal and it is very obvious that consumers are not spending like they did in 2024.

[BrewerFan]

14 hours ago, homer said:

I'd be curious to know how big an impact the advent of the internet had on economic numbers. AI might become a nothingburger when it's all said and done. Having said that, I'm glad I got in NVDA under $100. I absolutely think we're in some kind of bubble but impossible to say when it will burst.

I don't think this is a bubble... 

We saw it yesterday with META, MSFT and GOOGL all increasing their AI CapEx, you have these massive new deals popping up constantly. 

 

I think as the market caps grow, their valuation will start to come down, but AI CapEx is already going to be about 120B bigger next year to the ~480B with conservative estimates having it grow to 1T by 2030.

I don't see a world in which AI becomes a "nothingburger." 

FSD, Automation, just two area's in which AI could/should have a massive-massive impact on the economy. 

 

11 hours ago, nate82 said:

The fear of AI taking a lot of jobs is the same as the robotics fear.  Some jobs will be eliminated and some created.  Basic jobs will go away but more technical jobs and ones that need human input and interaction with won’t be as impacted.

This fear has come with every new revolutionary innovation. The internet, assembly line, etc...

They've all created more jobs.

Not entirely sure how AI does that, but...I suppose that's the point. 

[igor67]

I will say most of the AI tools I've tried have been very underwhelming. As in it is faster for me to just do the work. It can write middle of the road middle school level science and math questions, but I can just make those up on the spot. Anything beyond that and it couldn't, It certainly wasn't able to write Common Core level questions. I tried the features to have it try writing labs, and everything I got was pure hallucination which was readily apparent at least because the website sources it pointed to did not have info that supported the different labs. I found one that would OCR on older paper curriculum I had and convert it to Word that was actually moderately useful because it didn't give too much weird formatting  or mistakes handling exponents. But it wasn't enough to pay for after I used it a couple of times. So I'm mostly seeing signs of very limited markets. Similar to the great MOOC debacle a decade ago.

[owbc]

11 hours ago, BrewerFan said:

I don't think this is a bubble... 

We saw it yesterday with META, MSFT and GOOGL all increasing their AI CapEx, you have these massive new deals popping up constantly. 

 

I think as the market caps grow, their valuation will start to come down, but AI CapEx is already going to be about 120B bigger next year to the ~480B with conservative estimates having it grow to 1T by 2030.

I don't see a world in which AI becomes a "nothingburger." 

FSD, Automation, just two area's in which AI could/should have a massive-massive impact on the economy. 

 

This fear has come with every new revolutionary innovation. The internet, assembly line, etc...

They've all created more jobs.

Not entirely sure how AI does that, but...I suppose that's the point. 

The economy is pretty blah outside of the AI investment. What worries me is that there is a lot of shady stuff going on based on valuations that are assuming this stuff is going to eventually be making money. I think the insular deals between tech and AI startups will be ugly when they unravel. 

One example I saw was OpenAI’s recent foray into erotica. Does that sound like something a startup with a half-trillion valuation should be doing if the revenue projections were all rosy? 
 

I think AI tools will be great and have a lot of practical applications but it’s more of the financial dealings during this crazy boom that worry me as well as the complete lack of anyone in government pumping the breaks or preparing us for what’s to come.