Transcription
Same day that AMD makes a major deal with OpenAI, leading a lot of people to say this is peak bubble and peak circular, which we'll talk about exactly what's going on there. There's a note in Bloomberg about comparisons from today all the way back to 1929. See, in 1929, the Dow Jones fell 17% as we started entering the Great Depression. A lot of people are like, "Ah, 7%. Who cares?" The market was up double in just the year prior to that. And we're only up 73% on the S&P 500 since 2023. So, we're not up as much. And it took two years. So, you know, we're not we're not well, of course not.
But, of course, well, we fell 17% in 2029. That was just the beginning of a slow, drawn out death of the stock market. And the circular financing of what's happening today with the AMD deal has a lot of people scratching their heads wondering, "OMG." After all, in 1930, the Dow was down 33%. By 1931, it was down 50%, by 1932, it was down 80% from its peak. Imagine you're investing in the Q's today at, you know, 607 and it falls 80%. Well, 607 * 0.2 would bring you down to $121 on the cues. Just to make that relatable to some of the numbers that we're facing today, if you take the S&P 500, which is trading for 6,4 $6,741, multiply that by 0.2, you'd be at $1,348 on the S&P 500 if we had an 80% decline, just like what happened in 1929 to 1932.
Now, that ended up leading to an explosion of bank failures. You had these crazy tariffs that were instituted in 1930. Good thing we don't make stupid moves like that. We had a belief in new technologies like electrification, vacuum cleaners, movies, radio, cars. You had radio stocks going from a $150 per, you know, stock per share to $85 per share. You can look up RCA and those companies are dawn today. Uh, and of course back then you had a lot of fraud and insider trading. Like back then you had a lot of painting the tape. Fortunately today you don't have painting the tape because it's illegal. But I I guess insider trading some say is exactly still what happens given that the Kabesi letter says here on October 2nd users on X began noting unusual options activity in AMD. Over $6 million worth of long-dated call options were purchased and flows into AMD stock were strong. One user explicitly said they bet AMD just signed a multi-billion dollar deal to support GPUs to OpenAI. 4 days later, OpenAI and AMD announced a deal that will generate tens of billions of dollars in new revenue. Of course, I I thought the grifting was only possible in a certain property on Pennsylvania Avenue, not all over the stock market. But apparently, it's possible all over the stock market. And this is where we got to talk about the potential circular nature of these investments. And is there any risk to all of this? Well, let's figure it out.
So, here we go. Let's understand exactly how this all breaks down. So, I'm going to try to walk through this as simply as possible. So, the first thing we have to understand is that Nvidia just made a commitment that they will invest $10 billion into OpenAI for every gigawatt that is deployed. Okay, this is going to be a little hard to follow, so I'm just going to walk you through it here one step at a time. Nvidia will invest $10 billion once the first gigawatt is deployed. And then every next gigawatt they'll do another $10 billion dollars for a total of hundred billion dollars. Now when is the first gigawatt expected to be deployed? Well, OpenAI tells us. Here's the September 22nd press release. And if we scroll down right here, ah the first gigawatt of Nvidia systems will be deployed in the second half of 2026 on Nvidia's Vera Rubin platform. Ah, okay. Got it. All right. So let's write that down. So the first gigawatt will be deployed in the second half of 2026. Okay, got it. Now when that first gigawatt is deployed, what happens then? Well, then $10 billion of investment goes to OpenAI. Okay, so now OpenAI gets a bunch of new cash. And what is OpenAI now promising that they're going to do? Well, uh, in the second half, Open Door or Open Door, OpenAI is expecting to invest with AMD into a new deployment of at least 1 gigawatt of energy starting in the second half of 2026. That's because today we heard that there was a strategic partnership announced between AMD and OpenAI where they're going to deploy 6 gigawatts of AMD GPUs. Listen to this. You can't make this up. The first gigawatt deployment of AMD is set to begin in the second half of 2026. So let let me recap that for a moment. Okay, AMD is not in the picture and you got Nvidia and OpenAI. Nvidia says, "Hey, we'll invest $10 billion dollars into OpenAI when we deploy this new gigawatt expected to be done in the second half of 2026." Then OpenAI is like, "Okay, hey AMD, you want to start construction once we have this inflow of $10 billion from Nvidia into a gigafactory or, you know, a gigawatt factory? I'm thinking Tesla here. Uh, in the second half of 2026, we'll start that." It's like, "Yeah, let's do that." You see the circle building so far? And I'm not making this stuff up. These they're literally writing it here on the Nvidia OpenAI deal. We will deploy it in the second half. And on the AMD OpenAI deal, we will begin in the second half of 2026. It's literally like the AMD deal is waiting for the Nvidia deal to be done or progressed. So the money from Nvidia goes to OpenAI which can then go to AMD.
But it gets even more crazy than that because after that money gets deployed then AMD gets revenue right because AMD is going to sell chips to uh OpenAI but who also gets money well Nvidia is also getting money because their data centers are getting completed with open. both deals are going at the same time, right? But then it gets really circular because the more money Nvidia gets, the more money Nvidia can keep putting into OpenAI to do the second gigawatt, the third gigawatt, the fourth gigawatt, the fifth, all the way up to 10, right? What does AMD do when they get money? Well, when AMD gets money, AMD's stock price is going to go up. But what's crazy is OpenAI is going to have ownership in AMD. And if OpenAI has ownership in AMD and AMD stock goes up and Nvidia has ownership in OpenAI, then Nvidia directly benefits from AMD stock going up, which then means there's more money for Nvidia to do more deals, which if Nvidia does the second and third gigawatt of the more of more deals, then technically you could go back to AMD and you could go do the second gigawatt, the third gigawatt, the fourth gigawatt, and just always be like 6 months behind the Nvidia deal, just like they're blatantly telling you here again, deployed in the second half, starting in the second half, and then it'll be the second gigawatt, the third gigawatt, and the cycle is literally just a giant circle based on stock price at one point because of this AMD commitment because the AMD commitment Uh the deal is based on Nvidia hitting certain stock price targets which is wild because AMD all right this is something like let's back back off the circle for a moment. You need to understand this about AMD. AMD does not have as big of PP as Nvidia. That's just the reality. Nvidia has much larger PB. Let me prove it to you. on screen. AMD has a gross margin of 39.8%. Nvidia has a gross margin that's nearly twice as large, 72%. That's called big pricing power. But then when you look at the net income, AMD has net income this last year of 11.3%. So their net margin is 11.3%. Last year is only 4.5%. But even if we look at the 11.3%, it's still 1/5 of Nvidia's net income. Nvidia smokes AMD's pricing power because the margins are insane. The numbers don't lie. This isn't me judging someone else's PP. I don't do that. I don't judge other people. I just read the financials and the PP shows itself. I can measure the PP by going to sec.gov. It's really simple. So when we understand that Nvidia has really big pee pee here, why is AMD giving up 10% of their company? Because they have to. It's branding. Now Lisa Sue can tell everybody, "Hey, look, OpenAI, the largest AI company in the world, is betting big on AMD." They're giving up 10% of their business. But they're doing that because they want to use it as marketing to get more deals with other companies. Maybe Anthropic will be next or maybe Microsoft or what. Who knows? All these other companies will come around because of the branding. Some of this is evidenced by the fact that when Lisa Sue was on Bloomberg this morning, she was asked, "Oh, so where are these data centers going to be, Lisa Sue?" And she's like, "Oh, oh, where? Oh, well, the world will actually need data centers everywhere. So they will be in diversified locations to be determined. You can't make that stuff up. That's exactly what happened. And it's because you don't really have to make your plans yet at AMD because you're sort of kind of at this point it feels like contingent on the Nvidia buildout because once the Nvidia buildout happens, OpenAI gets $10 billion of cash. Now OpenAI doesn't really have cash. OpenAI is expecting revenues of about $12.7 billion this year, but they're also expecting to be a money loser between now and probably 2030, expecting to blow over a hundred billion of expenses themselves on either training AI or AI buildouts or whatever. And these are expectations we're getting reported by the information. So, a lot of these insider funding round leaks and more.
So, what stage of the cycle does all this kind of crap put us at? Well, we know this is a giant circle. Again, OpenAI gets investments from Nvidia to go build out factories to go buy Nvidia chips. The more money Nvidia gets, the more they can reinvest back into OpenAI. OpenAI signs deals following those investments with AMD, giving AMD shares based on the stock doing well. AMD gets those share or gets that, you know, uh or sorry, OpenAI. God, OpenAI, it's so circular. I trip over this. OpenAI gets the investment. AMD, which is worth more as the stock price of AMD goes up, which lets OpenAI have a greater valuation, which boosts Nvidia's valuation, which lets Nvidia go do more. I mean, you get it. It's a giant circle. So, what's what's the stage of the cycle? Well, here's usually how cycles work. When people are really fearful, the only people who buy when there's fear are people with cash. A lot of cash. A lot. Everybody buys with cash. And then after people do a lot of buying with cash, you get FOMO buying. And then after you get a lot of FOMO buying, you get FOMO meets debt. I kid you not. Those are the words that are being circulated right now like on Doomberg or otherwise is that you know what's keeping this going is debt. You know here you've got Meta for example turn to lenders to secure $26 billion in financing. If you want to see debt just go look at Oracle. Oracle is like chalk full of debt. Uh, I actually give credit to a company like Amazon for using less debt than some of these other companies because some of the things are kind of starting to get scary with just the amount of FOMO meets debt. And that's the way it works. That's how you kind of keep fueling the cycle. Does that mean we're at the end of the cycle? Not necessarily. Because if we go back over here, we can see that after debt, we get euphoria. But then we get to a new special phase of the cycle. The special phase of the cycle is this. Stock prices start getting pegged to more euphoria. Huh. Okay. Well, good thing that's not happening yet, is it? Oh, wait. Vesting of AMD ownership is tied to AMD achieving certain share price targets and to OpenAI achieving the technical and commercial milestones required to enable AMD deployments at scale. Ah interesting. So, this deal is contingent upon AMD going to the moon tripling up to $600 at the highest trunch, which is really similar to what you've seen with Elon's stock-based compensation or the Open Door plan that oh, Open Door is going to go to $30 and then we'll get all this money in stock. That's a phase of the cycle where you start tying money being made or deals being contingent upon the stock price moving to certain levels. And then of course you get the mania and the bubble. What happens after that? Well, that's for a different video. We don't want to talk about that. That's that'd be crazy. Nobody wants to talk about bubble. Nobody wants to talk about 1929. Although apparently there's like a new book out now on the 1929 Andrew Ross Sorkin slapped together. I wonder how much GPT he used on it. But uh I thought the article was fascinating. Bloomberg did a whole write up on it. And it is interesting because valuations today, they're not technically as high as what we saw in the dot bubble, which means there's still room to go. Here's the screeny of it right here. S&P valuation still below.com highs. This is crazy because I kind I hate comparing anything to the dot com era, but if you look at the Warren Buffett indicator, we are at the highest level we have ever been in valuations uh for the Warren Buffett indicator. We're like 213% or whatever as a percentage of GDP. It was very interesting if I Warren Buffett indicator current valuation current market valuation.com strongly overvalued Buffett indicator US stock market divided by annualized GDP 217%. The highest we have ever seen in history possibly driven by some of the madness of these circular deals but whatever. I do wonder what kind of escape hatches companies like Microsoft will end up having. You know, Microsoft has like a $19 billion deal with Nebas. And I wonder like what escape hatches the big companies have because usually they they have some. The big companies are usually able to negotiate escape hatches because they're big. And the small companies want to work with the big companies, so they're able to get and negotiate exit clauses where the small companies are like, "Well, we'd be dead anyway if the bubble pops, and so we may as well take the upside now and then cash out while you can." That's typically what happens in a bubble.
But anyway, it's worth noting that right now there are estimates that we will need $2 trillion of artificial intelligence revenue to fully fund the buildout by 2030 of artificial intelligence that markets are expecting. The problem is revenues are only expected to be $1.2 trillion, which is a 40% miss. And of course, my big fear is that much like in 1929 where we started, you know, bailing out companies, it won't be enough. We'll cut rates to zero. We'll start printing money. And it won't be enough to get people back to work because we won't have created AI jobs yet. We'll still be in an environment of not hiring people because of artificial intelligence. It's kind of scary, but it is what it is. Now, you know, in the short term, I mean, prices will move all over the place. I mean, this morning in the uh course member live stream, I argued that we're probably close to a top and the highest we would go is what I said would be 232 on AMD. We stopped out at about what 228 over here and we're talking the course member live like it's the the highest pricing is probably going to be in the pre-market. It almost always is. So, but that's in our course member live stream. So, anyway, welcome to the AI bubble.
>> Why not advertise these things that you told us here? I feel like nobody else knows about this.
>> We'll we'll try a little advertising and see how it goes.
>> Congratulations, man. You have done so much. People love you. People look up to you.
>> Kevin Pra there, financial analyst. and YouTuber Meet Kevin. Always great to get your take.