Transcription
Welcome back everyone. Today on the Joseph Carlson show, a lot has happened over the past week. Last week, the market was rattled with a rapid sell-off in the semiconductor ETF. It fell over 10% in a day and a half. South Korea's Cosby index cratered 8%. Tom Lee described this as a sobering event.
"And now, I think there's some sobering taking place because expectations are higher."
And then today, it seems like there's a bit of a reversal. Stocks are going back up. So, what is causing this hiccup in the market? What is causing this sobering event? Well, a lot of different factors, and it's about to get a lot crazier. We have a jam-packed week of macro events. Oracle is going to be reporting earnings. This stock is incredibly meaningful to the AI story. They are one of the ones selling all the cloud infrastructure. I'll be going over some expectations for this as well as Adobe that will also be reporting earnings later this week. We also have rumored news that Meta is weighing a big equity raise to finance AI infrastructure. We heard this story before, but it was with Google just a week ago. And I'll be going over why I support Google's equity raise and I do not support Meta's equity raise. We have a critical inflation report coming out this week as well, which many investors are concerned about that inflation could be creeping up. And while all of this is going on, we have big debates taking place in the market. One of them from the repeated AI skeptic and AI bear, which is Gary Marcus, tearing apart Google in a viral tweet saying that Google is building out a commodity, which is AI. We also have Steve Eisman reiterating the same thing that he believes all this AI spend is ultimately a commodity that won't have much pricing power.
"Difference between chat GPT, it's very commoditized."
"People are switching constantly from one to the other."
"There are no moats."
He doesn't even believe that AI has a moat at all. Yet, these companies are investing hundreds of billions, if not trillions of dollars. I am fully invested and fully exposed to the hyperscalers that are spending all this money. And I've heard this argument of AI being commoditized many times. So, I'll be giving today the full rebuttal of the AI commoditized argument. And then, of course, we have the fail of the week, which in this case is Sam Bankman-Fried, the imprisoned founder of FTX. Just today, we got news that Sam Bankman-Fried applied for a pardon from Trump. Now, a lot of you may have suspected this was happening. He is serving currently a 25-year prison sentence, but Sam Bankman-Fried may be one of the most unlucky prisoners in our entire prison system. We'll be going over all of it in the fail of the week.
Now, first of all, I have to mention that we haven't uploaded for a while on this channel, but if you want free content on a more regular basis, make sure you check out the Joseph Carlson After Hours channel. The majority of my new content has been on this channel for some time and it will continue to be in the future. Make sure that you're subscribed to the Joseph Carlson After Hours channel. The bell notifications turned on so you get notified whenever I upload.
Now, having said that, let's go ahead and jump in. Last week was an important week in the market, and the reason why is because it was the first week that we saw real weakness in the semiconductor trade. At the same time, South Korea's Cosby cratered 8%. It tripped multiple circuit breakers. And there's a couple reasons for this. First of all, like we mentioned, Broadcom reported their earnings. And investors were expecting the same thing that we've seen over and over again, a beat and a raise. You have to have that combination to have these incredible performances that we've seen. But Broadcom did not give investors a beat and a raise. They basically just beat and then they reiterated the same thing that they said last time. That's a change in pace. It's a deceleration from analyst expectations. Broadcom has been one of the biggest winners in the AI trade, yet over the past one-week period, it's down 17%. It's not a death blow by any means, but this wasn't good, and the market reacted accordingly.
We also had at the same time the strong jobs report, which is both good news and bad news. All of this trouble started happening Friday after the Labor Department put out a blockbuster jobs report, finding that the US economy added 172,000 jobs in May, more than double economist forecasts. They also revised upwards the number from previous two months. Basically, the job market isn't just holding up, it's booming. Investors had been banking on job growth slowing and the Federal Reserve cutting interest rates. But the report on Friday changed that equation. The markets are now pricing in rate hikes. A higher rate environment puts a damper on stocks, particularly those of the high growth AI companies. So, we have all of that going on rattling the markets last week. And then there's other factors playing into this. For example, some analysts believe that investors are pulling capital out of other companies to free up capital for the SpaceX IPO. The SpaceX IPO is massive. It's like $75 billion and that is going to have an impact on other companies. But I also think that SpaceX, um, $75 billion is a lot to ask the public to fund. So I think it is a big lift in terms of how you're going to fund that purchase. And so I think now when the stocks wobbled a little bit, I think people are raising cash.
While we can point blame at the Broadcom report or the jobs report or investors raising cash for SpaceX, there's also the fact that we had this news from Sam Altman admitting that AI costs are becoming a huge issue. The company's seeking to improve value as overspending becomes a meme. This is something that happened basically overnight. A lot of businesses found out that using AI is actually really expensive and many of them are pulling back. The Uber CEO Dara Khosrowshahi says that the company blew through its annual AI budget in a single quarter. He says that they're cutting back on AI spend, limiting the amount that each developer can spend on AI tokens. Prudence on AI spending was basically unheard of just a month ago. And now all of a sudden, many CEOs are saying the same thing. AI is becoming expensive. And all of this culminates to a market that's looking far more volatile. And unfortunately, it's going to get more volatile later this week.
Oracle's reporting earnings this week, and this is another important one. Oracle needs to not only post great headline numbers, which I'm sure they will, but they need to have those great headline numbers in addition to guidance. And in this case, the guidance is going to be far more important. Oracle needs to raise guidance. If they say that they're just reiterating guidance or they're just holding things steady, that'll be one more company that's showing a deceleration from analyst expectations. Another company reporting earnings this week is Adobe. Adobe will undoubtedly post great numbers, but they need to show some way that AI is not disrupting the business. They need to create a stronger defense for that. Right now, Adobe's looked at as a company that has temporary pricing power and a high probability of being disrupted by AI. It trades at a 10 forward PE ratio, a 10% free cash flow yield. It is by the numbers one of the cheapest high-quality companies in the market simply because of the theme of AI disruption. So Adobe management needs to work on convincing investors that AI is not disrupting the business by showing organic volume growth of products and seats. It's a hard story to tell and the market has pivoted directly against them.
Now out of all the volatility and themes that are going on today, there's one that I think is important to address and this is one that involves most of my core position. When I look at my portfolio, I'm heavily invested in artificial intelligence through four hyperscalers. We have Meta, Google, and Microsoft, all spending hundreds of billions of dollars in capex. And in my other portfolio, the Story Fund, we hold Amazon, which is another one spending hundreds of billions of dollars in AI capex. Between those companies, Meta, Microsoft, Amazon, and Google, we have the majority of AI spend, the companies funding this big capex buildout. And this is what's caused a lot of the most controversy over what type of returns this investment will get. And many investors believe that this level of spending is transforming these businesses into worse businesses.
"But if you're going to spend when it spent $90 billion last year,"
"Yeah,"
"it could handle it from its own cash flow. When it's spending $180 billion, it's got to raise capital. And not just capital, stock. And you know, there were rumors that Meta is going to raise capital. Microsoft is going to raise capital. The the whole tech sector is being transformed because because it's so light from an asset light business to an asset intense business."
He first notes the transformation that these companies are going from asset light to capex heavy making it so that they have less money to return to investors. Many investors, including Steve Eisman, are concerned about this. They believe that the businesses are transforming in a very negative way. But he also doesn't highlight this as the only concern. And and the other thing I would just point out is despite the incredible sums of money that are being spent, what's being produced in terms of LLMs and generative AI, which is great, is not really differentiable. There's not that much difference between chat GPT. It's very commoditized.
"People are switching constantly from one to the other."
"There are no moats."
So for all the money that's being spent, what's being created, it seems to me at this point is a commodity.
This is a very damning take from Steve Eisman. And he's been someone that has correctly identified weaknesses in businesses. But as he highlights here, Google, Meta, Microsoft, and Amazon are becoming more capex heavy. And while they're becoming far more capex heavy, they're also investing in AI, which he says ultimately is a commodity and has no moat. Steve Eisman seems to have been influenced by Gary Marcus, or at least he shares his opinion because Gary Marcus is a massive AI skeptic and bear. He also went on the Steve Eisman podcast sharing his bearish views. But he summarized them in a recent tweet. He said why things will eventually fall apart. One, everybody, even Google, seems to be treating AI as though it's some kind of winner take all competition like web search was in which Google was taking over 95%. But everybody is building essentially the same technical solution with essentially the same data. So there is no moat. If there is no moat, nobody's going to take 90% of the market with no clear winners. Nobody can charge monopoly prices. Instead, you get the price wars of commodity pricing, which means everybody will wind up overpaying compared to the modest profits they'll be able to make in an intensely competitive regime. This tweet was viewed over 1.1 million times with over 6,700 likes. It is viral because it strikes at the heart of this debate that these companies, including Google, are actually becoming lower quality and less moaty over time.
The first thing that's worth mentioning is that he is right in one aspect. AI models are becoming commoditized. For example, if you just look at how many AI models there are, there are dozens and dozens. Right now, there's actually over a hundred different models from all different companies that have all different performance benchmarks on all different subjects. And although some of them slightly outperform others, they're always leapfrogging each other as well. For example, one month it might be OpenAI's having the best model, the next month it might be Grok, the next month it might be Google's Gemini, and so on and so forth. The point here is that there's a lot of models to choose from, and I can see why investors are concerned about this. When we look at companies like ASML, they're really one of a kind. There's not a list of the top 10 EUV machines. Or if there were, all 10 of them would be from ASML. So Steve Eisman is arguing that because there's so many models, there's going to be no pricing power with this product. And even though I agree with the foundation of their concern that AI is a commodity, I do not agree that they'll have no pricing power. Because we can look at other examples of how this worked with other companies. For example, if we look at Amazon, one of the biggest ways that Amazon makes money is through Amazon Web Services. Amazon Web Services existed long before artificial intelligence, and it was an incredibly good business before AI. One of the biggest products of AWS is S3, otherwise known as storage. Basically, just storing things in the web. Yet AWS offering online storage grew like crazy. It was growing 30% plus year-over-year, every quarter growing growing to $60 plus billion even before AI in 2022 and while it was growing at this rate, AWS commanded operating margins of 30%. So how was AWS able to sell something which was storage at margins far higher than commodity priced margins? AWS was selling storage in a durable way with global availability, permissions, versioning, life cycle policies, integration, analytics, security, compliance, developer trust, as well as thousands of tools developed around the storage to be able to handle and manipulate and massage the data to be able to use it in all different ways. See, storage was the commodity, but everything that AWS did to package the storage, to make it distributable, to make it secure, to make it easy for their customers to use was not a commodity. There's only a few companies that were able to ever partially match the services that Amazon offered. Again, before AI even took place, before we had the first Gemini model publicly released, Google Cloud was growing quickly. It was growing year-over-year at a fast pace, mostly selling the same thing that Amazon Web Services was selling, mostly storage and IT infrastructure. But this shows an important thing. What makes a commodity have pricing power like storage is when you take that commodity and you put it in a distributable service with all of these features around that commodity. It elevates the commodity from something that doesn't have any pricing power to having substantial pricing power. But these companies are not the only example of companies that have taken a basic commodity and have made it worth far more. You can take the example of Spotify. Spotify is a reseller of licensed music. They don't actually even own the product that they're selling. The music labels own it. Spotify sells a product that is offered on multiple other services with identical content. You can go on to YouTube Premium. You can go on to Apple Music. You can go on to a dozen other streaming services and stream music. So Spotify is ultimately a commodity product. Yet this company has extracted massive pricing power and margins are increasing. In the past trailing 12 months, they've generated over $3 billion in net income. Their free cash flow is growing rapidly and is projected to continue growing far into the future. The underlying product of music may be a commodity, but the service that Spotify sells is all the advantages surrounding that commodity: the distribution, the algorithms, the rankings, the social networking within the app itself, the user interface, all of that is important to the experience of accessing that commodity. Netflix is another hybrid example of this. Some of the stuff that Netflix creates are not commodities because they're creating original content. But much of the value offered by Netflix is simply licensing other people's shows, putting them onto Netflix where you have better discovery, better recommendations, better user profiles, better displays of content. Overall, the app makes the value. Part of the reason people pay for Netflix is they know that it's going to work well, that they're going to be able to discover shows that they want to watch, and that they always have something fresh to watch. People don't really care whether or not it's only available on Netflix. They just want to be entertained. Another example of a commodity product ultimately earning far more than commodity pricing is Texas Roadhouse. We can look at the return on capital employed for Texas Roadhouse. This is very similar and almost exchangeable for the return on invested capital. All this does is encompass debt as well. But the ROIC of Texas Roadhouse is 17.7%. It continually hovers around 17 to 20%. For a commodity, the return on capital ranges from 8 to 12%. That's typically the returns you get for a fully commoditized product. Yet, Texas Roadhouse is getting returns on capital of 17 to 20%, around double what a commodity would earn. How does Texas Roadhouse earn double the returns on capital of a commodity product when what they're selling is steaks, which is an outright commodity? Everybody can sell steaks. There is no barriers to entry to selling steaks. You'll see a common theme here. Texas Roadhouse isn't really just selling the steak, the commoditized product. They're selling the service, the consistency, the experience. They're selling a good time that other people have a more difficult time commoditizing. Putting together all the operations around how to sell that steak, in what way to sell it, how to make it efficient and good and consistent for every guest is the difficult thing that they sell. And that's where they earn the above commoditized returns of 20% on returns of capital instead of eight. So Marcus and Steve are coming to the wrong conclusions here. They're saying because other people can build the product or that the product itself is a commodity that means that there is no moat. But that's certainly not true. That would be like saying that Spotify has no moat, that Texas Roadhouse has no moat, that Netflix has no moat, or that Amazon's AWS has no moat. All of these companies have taken something that is ultimately a commodity and wrapped it into an ecosystem and a foundation that makes it so that they can extract above commodity pricing. So it's not about who has the best model or how many models exist. It's about which companies actually have the distribution, the full stack integration, the entire foundation to distribute it in a safe, efficient way. Those companies will be able to extract far higher pricing as they'll aggregate much of the AI demand.
The other important thing to address here is the concern of these companies becoming lower quality, moving from capital efficient companies to capex heavy companies.
"The whole tech sector is being transformed because because it's so light from an asset light business to an asset intense business."
Are these companies really being transformed from asset light to asset heavy? Well, the answer to that question is yes. These companies are moving from a capex light business to a capex heavy business. And in basic financial theory, that is moving from a more attractive company to a less attractive company. Now that's something that I believed for quite a while. But over time, my thoughts have been shaped on this a little differently. I actually believe that in many cases, being capex heavy does not mean that the business is lower quality. There are a lot of companies that require far more capex and have lower margins that have historically proven themselves to be incredibly good investments and good companies. Costco being one of them. Costco's management repeatedly says on the earnings call that they are not a margin company. They don't focus on margin. They don't care about margin. They focus on topline revenue growth. They expand their business over and over again organically through opening up expensive warehouses and expensive properties, but making them so that they're highly efficient in how they operate. Because of that, they're able to extract a tiny amount of profit for the amount of revenue that they generate. Their profit margin is below 3%. Their operating margin is below 4%. Their gross margin is only 12%. This is by the numbers one of the lowest margin businesses that we can look at. Even though Costco operates as an ultra low margin company, they demand one of the highest multiples in the stock market. Today, it trades at a 50 trailing PE ratio and a 45 forward PE ratio. Investors are willing to pay 40 plus years for their earnings because they know that even though the company's low margin, it is consistent, it's reliable, and it will continue to generate excess returns for long into the future. So being capex heavy, even lower margin, is not automatically a bad thing and it doesn't mean that they can't generate high returns on capital. And we'll see what these companies do over the next year. But I remain optimistic that I think that they are going to have superior returns with their distribution, with their infrastructure. They should be able to find many ways to monetize a product even if that underlying product is a commodity.
Now, on this note, we have the news that Meta is weighing a big equity raise. This is similar to what we heard with Google. In fact, it looks a little bit like Meta is just copying Google. They've seen that Google is going to get a lot of money by diluting shareholders by like 1.8%. And Meta is saying, maybe we can do the same thing. Now, I hold both of these companies, Meta and Google. And I believe in the case of Google, this dilution is a good thing. But I do not like it for Meta. And I hope that they don't end up doing this for a couple of reasons. There's two different reasons why I support this with Google and I think it's a good strategy for them, but I don't with Meta. And the first one is valuation. When we look at the forward PE of Google, it's currently sitting right around 30. So Google's at a 30 forward PE. When we look at Meta's, it's right now at a 19. Meta is at a way, way lower valuation, which means that when you dilute shareholders, you're doing it at a much more substantial cost than when Google's doing this. For example, in order for Google to get $80 billion, they need to dilute around 1.8%. For Meta to get the same $80 billion, they'd have to dilute around 5 to 6%. So, you're talking about a substantial level of dilution. And even diluting any amount when the stock is this cheap does not seem like a smart move. If Meta traded up to a 30 PE ratio, I would have no problem with them diluting. I think that it would actually make sense in that case, but at a 19, it just doesn't make sense. The next thing is that I just believe that Meta investors will react far more poorly to this news than Google. Not only because the dilution would be far more expensive, but because right now investors in Meta are a little bit more concerned about the return on capital than Google. Google has somewhat proven that they're going to have decent returns on capital, or at least investors believe that. But with Meta, that's still a big question mark. So, I think this could actually hit the stock price in a far more negative way for Meta than it did for Google. So, even though this is just rumored news, I'm hoping that Meta finds a different way of raising capital than diluting shareholders.
Now, in today's fail of the week, we have Sam Bankman-Fried of FTX. The FTX co-founder Sam Bankman-Fried has officially filed for a pardon from President Trump. Bankman-Fried is serving a 25-year prison sentence after being convicted of fraud and money laundering charges in 2024. We already know that President Trump likes pardoning people. For whatever reason, he's pardoned hundreds of people in his second term. And Sam Bankman-Fried is believing there may be a shot. It might be a long shot, but there is a shot after all that Trump pardoned Sam Bankman-Fried. After all, Trump has pardoned fraudsters in the past. Trevor Milton, the founder of Nikola, was convicted of defrauding investors. There's much evidence showing that his presentations and his statements about the company were not factual, but he's since been pardoned by President Trump. Now, Sam Bankman-Fried believes that there may be a chance, and now he believes it's the right time to strike. He might be able to get it in Trump's head that he is worthy of a pardon, that he's an individual not deserving of a 25-year sentence.
When I look at Sam Bankman-Fried's case, I have mixed feelings on this because first of all, he did dig his own grave to a certain extent. Sam Bankman-Fried did defraud investors. FTX customers deposited their money into crypto believing it was being held for them or used in an authorized way. Instead, prosecutors proved that Sam Bankman-Fried was secretly moving that money and diverting it to his own trading firm. Then he was using it to trade on different venture capital ideas. He was trading on real estate. He was trading on loans to executives. He was making political donations to it. All of these things are fraudulent. And what Sam Bankman-Fried did was lie about the ways that he was using other people's money. That's illegal. And usually that ends up with jail time. But I also consider Sam Bankman-Fried to be one of the unluckiest people in prison today. First of all, one of the investments that Sam Bankman-Fried made with that money that he was illegally trading, he invested $500 million into Anthropic, buying 7.84% of the company. To give you an idea of what that investment is actually worth in current day, it's likely worth upwards of a hundred billion. So his investment that he originally made with the money illegally ended up being one of the best venture investments ever. He turned a $500 million likely into a $70 to $100 billion investment. This would be one of the greatest tech investments in history. The only problem was he did it with the wrong money and he did it at the wrong time. Sam Bankman-Fried was very unlucky that the crypto crash happened before the full value of Anthropic was realized or it became at least obvious. The timing of this was most unfortunate for Sam Bankman-Fried. If you think about it this way, if the crypto crash happened only a number of months later, FTX would have never been insolvent. They would have been able to afford paying out every single customer simply from their venture capital investments like Anthropic alone. That investment alone would have funded every single account, including any type of gains. It was just a matter of timing. Sam Bankman-Fried is a victim of his own stupidity and using money in the wrong way. But he's also a victim of bad timing. Had this all worked out just a couple months later, he wouldn't be in jail for 25 years. Another thing that Sam Bankman-Fried is uniquely unlucky for is the fact that FTX is one of the largest frauds in history where all of the people that were involved in the fraud are expected to get back not only their entire investment but also interest of 18 to 20%. So he has a massive fraud where everybody's getting repaid in full plus interest, which is unique. In some cases, even with massive frauds like Bernie Madoff, they can recover a lot of the money. They can get 90 plus% of the money back, but rarely do they get 100% of the money back and even more rarely in a fraudulent case do they get beyond 100% back. That is incredibly rare. So Sam Bankman-Fried returned investors' money plus interest and is facing 25 years in prison. That is why he is the fail of the week. That's all for this episode. Hope you enjoyed.