Section Insights
Current State of the AI Market
What is happening in the AI trade and tech market right now?
The AI trade appears to be struggling despite positive earnings reports from major companies. The market is currently headline-driven, leading to negative stock reactions even after good results. Major semiconductor stocks are down, and there are concerns about competition affecting companies like Nvidia.
- The AI trade is experiencing a downturn despite positive earnings.
- Market reactions are driven more by sentiment than fundamentals.
- Semiconductor stocks are particularly affected by competition fears.
Market Corrections and Future Opportunities
Why are stocks down, and what does it mean for future investments?
Stocks have corrected after being overvalued, creating potential buying opportunities. Factors like midterm elections, inflation, and year-end tax strategies are influencing market behavior. Despite current downturns, many strong companies continue to grow.
- Market corrections can present buying opportunities.
- External factors like elections and inflation impact stock performance.
- Strong companies like Amazon and Google are still performing well.
Long-Term Growth Prospects for Major Tech Companies
What are the long-term prospects for companies like Amazon and Google?
Companies like Amazon, Google, and Microsoft are expected to grow significantly over the next 5 to 10 years. Amazon's AWS division is seeing substantial growth, particularly in AI-related services, and the custom chips business is also expanding rapidly.
- Major tech companies are likely to grow and improve over the long term.
- AWS is experiencing strong growth, especially in AI revenue.
- Custom chip revenue is increasing significantly for companies like Amazon.
Reddit's New User Policy and Its Implications
How will Reddit's new user policy affect its traffic and advertising?
Reddit's decision to require users to create accounts may improve targeted advertising but could also lead to a loss of traffic if not managed well. The success of this policy depends on whether Reddit can maintain its relevance and traffic amidst competition.
- Reddit's new policy aims to enhance user targeting for advertisers.
- The effectiveness of the policy will determine Reddit's traffic and revenue.
- Maintaining user engagement is crucial for Reddit's future success.
Challenges Facing the Semiconductor Industry
What challenges are impacting the semiconductor industry and companies like AMD and Intel?
The semiconductor industry is facing challenges such as high memory costs, inventory issues, and supply constraints. Intel's recent guidance indicates struggles that may impact AMD as well, but AMD is positioned better due to its cash flow and faster growth in the data center sector.
- The semiconductor industry is experiencing significant challenges.
- Intel's struggles may affect AMD, but AMD has stronger cash flow.
- Data center growth is a key area for semiconductor companies.
Transcript
0:00 Is the AI trade over? Seems like every other day new headlines, company reports earnings, good comments, good results, stock react negatively. We have red across the board in the AI trade, semiconductors, data center providers, AI cloud players. You name it. So, what the heck is going on right now in the market? We had Amazon report, we had Intel report, we have Reddit. We'll cover all of that, of course, in today's video and I'll share my thoughts on what's going on at the moment. And this is just me repeating what I've been saying in the past as well. It is a headline-driven market right now. I've also said that this earning season is probably going to be one for the history books. We're going to see a lot of beats across the board, not all of them, but a lot. And we're going to see the market go red. I mean, as of right now, this is exactly what's happening. Now, does this mean that the trade is over? We've talked a lot about this during the live stream yesterday.
0:59 Is the trade over? Why would people buy semiconductors right now since the stock is flat for the last couple of months? Well, yes, the stocks or most of the semiconductor names, the bottleneck names, the new clouds, yes, they're flat over the last couple of months or down. But a lot of them, and maybe actually I'll show you with the heat map. This is what we're looking at right now pre-market. Very, very ugly in the tech space. Nvidia is down 3%, Broadcom is down 4. This also never made sense to me. If Nvidia is down on the fears of more competition, the competition is Broadcom. The competition is Google, it's Microsoft, it's everybody that's using Broadcom to make their own chips.
1:38 But hey, the market thinks actually let's bring everyone down. Okay. Micron is down close to 6%, AMD 8.6. That's because Intel Intel is down 24% pre-market, I believe. And Reddit is down close to 20% as well right now. But when we go and have a look at the year-to-date performance. This is my point. Year to date. All of these names here, the bottom left, all of these names are up tremendously. Now, I know not everybody bought at the start of the year, but yes, I do assume that a lot of people did buy at the start of the year.
2:14 Otherwise, the stocks would not have been up so much up until now. Of course, there are some other names, very, very big names, very good companies, a Meta, a Google, a Microsoft. Those names are not up for actually two of them are down this year, but they're underperforming the market. But my point here is that a lot of these names, the stock has just outrun the business. And right now, it is the business's job to catch up to valuation. Of course, we would all like to see all of these names go up every single day until we decide that we want to retire. But it doesn't work like this. Like an AMD, AMD is still up 158% year to date.
2:54 Why? Because AMD is expected to see its business accelerate from here, but the business needs to catch up to the valuation. It's the same with a lot of the other companies that we're seeing that are up triple digits year to date. I mean, look look at look at the cybersecurity names. They're still up. Fortinet is still up 74% year to date despite crashing yesterday. CrowdStrike, still up 120% year to date. Palo Alto, 105% year to date. And by the way, Rubrik is also up a lot year to date. And I'm actually thinking of selling completely out of Rubrik because the valuation to me right now doesn't make sense when I get to buy the other companies at much cheaper valuations without having to think about, oh, what happens when Rubrik reports and that stock might crash 20, 30% or so in one day. So, yes, as of right now, it seems like the narrative has switched completely to, oh, AI trade is completely over. There's no growth. There's But there is. When you look at each and every company, every one of those companies is telling the same thing. We are growing. We're growing as fast as we can. If we could magically have more supply, we would be growing more and more and the opportunity is immense, which is why we are investing right now. That doesn't mean that the stocks cannot go down. No, an Intel Intel went up, yes, based on a small turnaround that is happening inside the company, but it went up way more than that small turnaround. It was worth over half a trillion dollars at one point. Of course, if we do get suddenly announcements for foundry huge customers for Intel foundry, then yes, here again, the sentiment will change, but the business profile will change as well, because suddenly you have a backlog that's going to grow substantially. But a lot of the names that have gone up quite a lot and are down quite a lot have gone up too much based on vibes, based on yes, sentiment, some of them based on growth, but they overdid it.
4:58 And now they corrected. Valuation is catching up to the stock and I think that's when the opportunities are there for people that would like to buy more. Now, to the non-stock related part, midterm elections, Iran, those are still two things that are going to happen pretty soon. So, that puts some pressure on the stock market. It's also the end of the year. Maybe some tax loss harvesting can happen right now. That's always the same excuse every single year. But you have to take all of that into consideration.
5:32 Historically, a midterm year is not always the greatest one. Post midterms looks a bit better, but of course, this time might be different. Right? We have still inflation. Well, we just had a rate hike a month or so ago. So, it's a bit different, but you have to take all of that into consideration. Not only look at oh, the stocks are down, story is over. No. Story is not over for the majority of the big good names.
5:57 If you look at the performance, you'll see Amazon still performing very well. We talked about Google, Microsoft, Meta, semiconductor names. They're still growing and growing very, very quickly. Now, Intel had other issues, which we'll talk about that as well. Might impact AMD, which report, I believe, next week. And so, let's start off here with Amazon. Amazon is now flat year-to-date. And look, Amazon had a good report, but I'll have to repeat myself again. It doesn't matter if you had 1,000 hours of research in Amazon and you played the weekly options. Huh, look at what happened.
6:32 Stock is down 6.5%. Look at what happened with some other names as well. Great report, stock is down. Don't try to trade on the weekly stuff because we don't know what the market's mood is going to be. But Amazon, Amazon, yes, it is getting again to a stage where it it is extremely cheap. It is extremely cheap for this type of business. Of course, the main main number, the topic here, AWS. AWS, yet again, increased 39.3% year-over-year to close to $46 billion in revenue.
7:06 Incredible stuff. Margin did drop a little bit though, but not by much. If we look at advertising, advertising did see quite a big drop, actually, in growth rates. Only grew close to 11% year-over-year. So, was down to flat quarter-over-quarter, but there is a reason for that. And if we go and have a look at the custom metrics, which, of course, you can all try on fiscal.ai. There's a link down in the description and in the pinned comment, you'll get 15% off with my link and you new users will get fiscal pro for free for the first 2 weeks. Custom metrics means that you can do things like this or what we've seen with Google, search as a percent of revenue versus the rest of the business. That's how we realized that, hm, search as a percent of the overall revenue is now less than 50%.
7:52 And so, if we look back here at Amazon, AWS margins, 38.7%. Still very, very good, but down slightly from the 39.4% we've seen last quarter. Now, they also told us that over the next 5 to 10 years, they expect to have two businesses that will be worth a trillion dollars or will generate a trillion dollars. I mean, if we look at how fast this business is going to reach a trillion dollars in annual revenue, we should be looking at over $1 trillion in revenue generated in fiscal 28.
8:27 It's not that far from here. Now, if we look at the beats and misses for the quarter, you can see yes, a lot of misses here, but small ones, I would say. Revenue did miss slightly, diluted EPS missed slightly, but operating income beat, AWS sales beat as well, North America sales did miss a bit, international as well, advertising, subscription did beat, CAPEX was above the estimate, so unlike a Google and Microsoft, here CAPEX growth was a bit above that.
8:57 But look at this, 13.9% record operating margin for this business. We've been talking about this time and time again. Of course, free cash flow is a very, very important metric, but as of right now, we know why free cash flow is negative for these businesses. That's why you have to look at the core business. How profitable is the core business? And clearly, the core business is becoming better and better and more efficient. AWS is now at a $184 billion annualized revenue run rate. They are also expecting They're also expecting to grow AWS by 40% year-over-year in 2027.
9:35 And then another 40% in 2028, actually over 40%. And look at the size of this business already today. Now, I know it might not be the next multi-bagger or whatever, but it is one of the best businesses out there. This is one of these names, and actually, you can say the same for a Google, a Meta, a Microsoft. It's one of those names where you just buy and hold, and you sleep well at night. You know that this company will be a bigger company, a better company 5, 10, 20 years down the line. Okay, you know what? Maybe not 20 years we don't know, but 5 years down the line, 10 years down the line, yes, I'm pretty certain that these companies will be better and much bigger. During the call they told us the following thing, AWS backlog is not 578 billion dollars that grew triple digits year over year. Over 50% of the AWS AI revenue comes from inference. They've seen 7x increase of Fortune 500 spent on bedrock year over year and this is the sixth straight quarter of AWS acceleration. Now, what happened with advertising? Why did advertising growth slow down that much? They told us that Prime they shift cut around 850 basis points from Q3 at growth and beyond Prime boosted Q2 and resumes in Q4.
10:48 They're also lapping Q3 2025 search marketing change and we should be expecting to see heightened growth back again in the next quarter. Lastly, and that's the custom chips business, just like the TPUs for Google. Here as well it is growing and growing quite rapidly. Over 100% growth in the chips business revenue year over year. Over 20x Graviton and Tranium revenue commitments year over year. 30 to 40% Graviton price performance edge versus x86. And they also told us that they are going to start with external sales in late 2027.
11:24 So, just like with the TPUs, although TPUs are already starting to ramp at the end of this year and really ramping up at the start of 27, we're now going to have even more competition and more companies that could deliver their own chips to the broader markets. And so why is Amazon down after hours? Well, it's certainly not because it's a bad business. It's just the market's mood as of right now. Oh, more spending. Oh, less free cash flow. But if you actually look at the business itself, just like with Google, just like with Meta, I just don't see anything wrong with it.
11:59 Moving on to a company that well, unfortunately, there are some things that are not going their way at the moment and that's Reddit. Now, Reddit is an interesting one because as you can see, beats across the board. Beats across the board, only a small miss here in international revenue. But, as we've said, there are some other metrics that are more important, and that's the user metric. And user metrics across the board missed. Daily active users quarterly missed, weekly missed, US internationally, they all missed. Now, what did beat is ARPU, ARPU globally and US one as well. They did also tell us that logged in daily actives grew 7% year over year while US logged out daily actives fell 4%. The total was flat quarter over quarter, and that's the problem. This is great to see that they monetized their base. Great.
12:56 This is what we expected. But, what we wanted to see is maybe maybe they found a way to get more users on the platform. Because as of right now, okay, you have your user base, you can monetize them because monetization has been very very low and non-existent. But, for the future for the future, you would like your user base to grow more and more, so the monetization well, the big chunk there, the cash amount can continue to grow, grow, and grow. And so, the positive thing is that yes, ARPU grows quite a lot. They saw over 2x video ad revenue growth year over year, 1.3x revenue growth versus adjusted cost growth, ARPU grew 36% year over year, they seen an adjusted EBITDA margin of close to 45%, but the comments for next quarter are already not that great. They are now expecting to generate over a billion dollars in revenue, but that growth implies 40% year over year. So, again, a decline, a deceleration in growth rates.
13:54 They also told us that traffic fell meaningfully in Q3, and the decline accelerated late in the quarter. October is down more than 1 million referrals per week. If the trend holds, daily actives could fall by a low double-digit percent from the 130 million base. And to tackle the loss of traffic because of well, Google and the other chatbots out there, they're going to fully test in Q3 logged in only in select non-US market to make sure that they retain the users and protect the data. Meaning if you want to go on Reddit, you'll have to be a signed in user. This is something that we've seen on X. This is something that we've seen on Facebook, on Instagram, on LinkedIn as well. And this is I think what happened with Pinterest at the start. So, this is not something new.
14:43 The question is is it going to work? Because as of right now, yes, AI overviews, AI modes is a huge success. We've heard that from Google, but that sends very little traffic to Reddit. And this is something that we've talked about before. If Reddit can make a deal with Google, with Anthropic, with OpenAI, then that's great. But if they cannot and they're still losing traffic, that is a very very bad scenario. Now, if their test is successful and if you want to go on Reddit, you'll have to create an account. You'll have to be logged in. By the way, logged in means that advertisers, they can target those users way better than logged out users, which of course is better for Reddit because then average revenue per user continues to go up. But this could also mean that oh, maybe Google will now have issues to provide good answers to whoever searches for something because Reddit is still is still in the top 10 most visited websites on the planet. And according to them, Reddit is a top Google search term, meaning people search something and then they add the Reddit word as well. So, they go directly to that answer.
15:52 So, it's up to Reddit to make sure that look, if they say that their data is extremely valuable, they should be getting paid for it. If they do not get paid for it, then I guess the defensive move that they're doing right now is the right move to do. They're going to test it out and see what happens. It's definitely going to help monetization, but could potentially hurt of course the user experience for the logged out users. But then again, just sign in once and you don't have to go through this same process time and time again. Lastly, here with Reddit before we jump into Intel, Reddit is still a company that's expected to grow quite a lot. Right so right now for fiscal 26, we're expected to see $3.3 billion in revenue, 4.4 in fiscal 27, and then $5.5 billion in fiscal 28. So, a lot of growth is still ahead. Not only that, they're going to generate more and more free cash flow over the next couple of years. And if if the stock does remain cheap and they can continue to generate the amount of free cash flow, then yes, they'll just buy back more shares. They still have I think over $600 million worth in buybacks to deploy.
16:59 I sure hope that they're going to accelerate the buybacks now that the stock is back at $100 per share or so. At this valuation, this margin profile, still with the growth rates that they're experiencing right now, I do think that they should buy back more shares. Now, of course, for the long run, for the long run you would have to see the user base grow. And so for me it's quite clear to me Reddit right now, I'll put it again on the side. I'll wait and see what happens with the traffic, with the business itself. Couple of catalysts, maybe an AI data licensing deal. My position is small.
17:36 It has remained small for a good reason. The reason is look at the reaction right now. Then if we do get some data from external sources saying that oh, actually traffic is increasing, then I'll probably buy more shares. Last but not least, and that's Intel. Intel is still up 75% year to date, but now experiencing a drawdown of 51% from the peak. Now, Intel is a $360 billion company right now. And if you look at this, this is what has happened with Intel. Of course, the reason why the stock jumped so much is because we've started to see okay, finally this company will start to grow a bit more.
18:17 And that's was the turnaround effect. Now, what about free cash flow? They already told us something during the call with regards to free cash flow, but let's have a look at what the market is expecting. The market is expecting free cash flow to only inflect in fiscal 28. Management told us that in fiscal 27, it's probably still going to be negative free cash flow, but we are expected to see a big big inflection point in free cash flow in fiscal 28. Of course, this number could be much much bigger if we do see foundry attracting a lot more clients. Now, let's dive a little bit deeper. So, for the quarter, we have beats across the board for the quarter and revenue that came in in line. That's also up 21% year-over-year, but the problem probably is next quarter.
19:04 Next quarter, big misses in Q4 revenue, non-GAAP gross margin, and adjusted EPS all missed by quite a lot. What we've seen for this quarter is that data center, which is bar right here, was up year-over-year quarter-over-quarter. That's great. The consumer part, and this is basically the struggle right now because well, memory is very expensive. You have inventory clean-up right now, not a lot of demand, but foundry foundry did add to the growth here. Operating margin for foundry is also getting a little bit better. Well, actually not a little bit, much better. Data center operating margin is at 41.5% up from 23.4% a year ago, but the big hit is because of the Q4 guidance. They did say that the industry-wide memory shortages are weighing on the PC market, which is also why we've seen AMD drop after hours.
19:57 Although again, AMD is a company that does generate a lot more free cash flow than Intel. The data center business is growing much faster than Intel as well, so it's it's a bit different than Intel. Now, a couple of reasons behind the Q4 guide. So, they shut down a fab in Israel about $300 million of Q4 revenue and 200 basis points of gross margin hit because of that. PC inventory flush, we just talked about it. And servers capped by supply.
20:27 So, data center AI Q4 revenue to grow mid-teens quarter over quarter. The first half of '27 is tighter on wafer and substrate, which goes back to what we've said in previous videos. People think that '27 is going to be a bad year. I think these companies, chipmakers, all of that, they're still going to grow as fast as they can. And they told us industry stays supply constrained into 2028. Now, as for the CPU side of things, and this is again something that AMD has told us before, they see over $200 CPU server CPU time in 5 years. That's up from $40 billion.
21:03 They see over 30% industry server CPU unit growth in '26 and '27. Around $10 billion 2026 data center AI revenue, which is above the start of the year plan. Three times expected purpose-built silicon growth over 3 years. And $20 billion in gross capex for '26, but '27 is expected to be up strongly. They've seen a 40% reduction in cost for 18A since Q1. Now, the increase in capex could be a sign that foundry is getting some attention.
21:39 We've heard from Li Bu at the start that they're not going to invest a lot if there is no demand. So, it's it's the opposite of build and they'll come. For Intel, it's it's not the same. It's come to us first and then and then we'll spend we'll spend the money. Which again, for the company that they are or they have been, I think that was the right move. As of right now, we can see that the cost for foundry is falling fast, which is very good.
22:05 Because if foundry does pick up a lot or at the start a few big customers, this is of course going to be a huge huge and profitable business for them. Now, with Intel, unlike Reddit and Amazon and a Google and a Meta, of course, it was very expensive going into the print. And there still is a a of or there still are a lot of questions going forward because yes, a lot of the future of this company, in my opinion, does rely on a successful Intel foundry. But Amazon, Meta, or Google, these types of companies have done very, very well, and it does seem like maybe maybe going to 2027, picking up the big guys, the big names, might be the right move. Or or it seems like the market tells you, "Just drop AI and go into some other places, Eli Lilly, financials, energy, things like that."
22:57 Maybe. But I'll I'll stick with what I have as of right now until the business proves that maybe it cannot grow into the valuation or it cannot grow as fast as what I'm expecting it to grow, which as of right now, nobody, none of the companies that I own, tell me that I should drop them purely because the business is broken. Now, there are certain companies that I would like to add more to. Some companies where I just said with Reddit, for example, I'll say, "You know what?
23:27 Cool. It's cheap. Well, maybe it's cheap for a reason. We'll put it to the side." As for a Rubrik, like I said before, it is very expensive, and seeing the other names take a big hit, valuation is way more attractive. Maybe I'll say, "You know what? Thank you. I like Rubrik, but maybe I'll come back when the valuation makes sense again." That's about it for me in today's video. Relax. We've seen this happen before. It's probably going to happen again in the future. Relax. Know what you own, and let's have a great end of the week and a great weekend as well. Bye-bye.
24:00 >>
Summary
- The AI trade is facing a downturn, with stocks like Nvidia and Intel seeing significant declines despite strong earnings reports.
- Many tech stocks have outperformed year-to-date but are now correcting as valuations catch up to business performance.
- Amazon's AWS continues to grow rapidly, but its stock fell due to market sentiment rather than business fundamentals.
- Reddit reported mixed results, with user growth stagnating, raising concerns about future monetization and user engagement.
- Intel's stock is down significantly following a disappointing Q4 guidance, despite a year-to-date increase; the company faces challenges in the PC market and memory shortages.
- Overall, the market is headline-driven, and external factors like midterm elections and inflation are contributing to stock volatility.
- Investors are advised to focus on long-term growth potential rather than short-term market reactions.
Questions Answered
What is happening in the AI trade and tech market right now?
The AI trade appears to be struggling despite positive earnings reports from major companies. The market is currently headline-driven, leading to negative stock reactions even after good results. Major semiconductor stocks are down, and there are concerns about competition affecting companies like Nvidia.
Why are stocks down, and what does it mean for future investments?
Stocks have corrected after being overvalued, creating potential buying opportunities. Factors like midterm elections, inflation, and year-end tax strategies are influencing market behavior. Despite current downturns, many strong companies continue to grow.
What are the long-term prospects for companies like Amazon and Google?
Companies like Amazon, Google, and Microsoft are expected to grow significantly over the next 5 to 10 years. Amazon's AWS division is seeing substantial growth, particularly in AI-related services, and the custom chips business is also expanding rapidly.
How will Reddit's new user policy affect its traffic and advertising?
Reddit's decision to require users to create accounts may improve targeted advertising but could also lead to a loss of traffic if not managed well. The success of this policy depends on whether Reddit can maintain its relevance and traffic amidst competition.
What challenges are impacting the semiconductor industry and companies like AMD and Intel?
The semiconductor industry is facing challenges such as high memory costs, inventory issues, and supply constraints. Intel's recent guidance indicates struggles that may impact AMD as well, but AMD is positioned better due to its cash flow and faster growth in the data center sector.