Tesla + SpaceX: If Not Merger, What Else?

Kind: captions Language: en A merger between Tesla and SpaceX is not going to happen. Something much better is going to happen. And I'm going to show you what it is. They're all telling you that a merger is inevitable between Tesla and SpaceX. Here's Chimath. >> The chances that Tesla and SpaceX could wind up being the same company being 100% is what you're putting it on. >> Let me be clear. 99.999%. >> Larry Goldberg, Tesla Boomer Mama. A true merger of equals. Bloomberg inevitable. It's not going to be a merger. It's going to be a Terraab joint venture. Terraab is going to be built as a joint venture between Tesla and SpaceX. The merger idea makes no sense. And I'm going to explain to you why. Because the merger idea does not generate additional capital to help build this monster. The terapab that they're going to build is going to cost over $50 billion for phase 1. They don't have it. SpaceX will raise $75 billion in the IPO. 50 billion of that is going to go to a new data center. They told us in the IPO documents that the first priority for the money raised is for terrestrial data centers. Tesla has 44 billion in cash. They have all kinds of money going into Optimus, going into Robo Taxi. They are very capex heavy for this year, next year. They don't have the free cash to put up $25 billion or $30 billion for Tariff app. Neither one of them has enough cash to get all the way to what's needed. But a joint venture can raise that money and make this work for both companies and for outside investors who add sufficient capital to make this work. That's the secret. That's the difference between a merger and a joint venture. And we know, we've already been told that it's a joint venture. And everyone's ignoring the fact that we've been told it multiple times. Bloomberg and reporting on Tarap has multiple times said staff working for the joint venture between Tesla and SpaceX. People are interpreting the term joint venture as if it just means oh they're working together. It's a partnership. It's a team. No, a joint venture is forming a company. I don't think they've formed the company yet, but they're maybe they have. TF Terafab Inc. probably not the logo. I'm going to show you how this works. I'm going to show you why this makes sense. I'm going to suggest to you how this might work. I'm not saying it'd be precisely like this, but this is how it might work. You can imagine here Tesla puts in $15 billion. SpaceX puts in $15 billion, and they find outside investors, not like through an IPO, they find venture funding, some sort of private equity funding, $20 billion. Maybe Nvidia chips in, maybe Intel chips, and Intel's already supposedly a partner in this project. So you get a total of $50 billion. SpaceX owns 30%, Tesla owns 30%, the outside investors own 40%. And that funds Terraab Inc. at $50 billion. I have another theory that Elon chips in $5 billion of his own cash and that makes it 55 billion, which is precisely the number in the filing with Grimes County for how much the first phase is going to cost. a theory maybe Elon throws in an extra 5 billion to show his commitment that helps reassure outside investors that he's in. I think Elon probably becomes CEO or runs the joint venture himself. I don't think that's none of this is necessary. I just think that part makes sense. But this is a funding example. And then when they get ready for phase two, which is another $55 billion, then assuming each company has now succeeded in its businesses, Robo Taxi is running well, maybe Optimus is generating revenue, Starlink has grown, and the companies have more free cash based on on generating revenue as opposed to outside funding. and they chip in another 15 billion a piece and they get outside investors to chip in another 20 billion or they just fund it themselves. If they have enough free cash, they just fund it themselves. But phase one, this works for phase one. This is enough cash to fund phase one. Now, how does this work mechanically? Why is this good for both companies and for outside investors? Imagine a scenario. I'm not saying this is the actual cost, but this is just simple numbers. Suppose it costs them $500 to make a chip. This is an image of the AI5 chip that in theory will be made in terapab. I think they'll really be making AI6. I think it may cost more than this. It might really be a,000 and 2,000. It might really be 2,000 and 4,000. First rule is the purpose, the main purpose of building Terraab is to have sufficient supply of chips for Tesla and for SpaceX for what they want to do. So Tesla and SpaceX get priority. If they if there's chips available, Tesla and SpaceX get first crack at any chips. Outside buyers have to wait in line. If there's no chips available after Tesla and SpaceX get their piece, there's no chips available. As far as we know, there isn't a lot of demand from outside for these chips yet. But if they become very efficient, if they're very good at inference, as we expect they will be, there might be outside demand, and that could be a big source of revenue and profit for the joint venture in the future. And remember Tesla and SpaceX will own together 60% of this company. So if it makes a profit, 60% of the profit goes to Tesla and SpaceX. That's still very good. Now if Tesla or SpaceX buys a chip, instead of paying the $1,000 market value, they only pay $700. There's a net profit. This is net, not gross. Net revenue, net cost, net profit. I realize that that's this is playing with numbers a little bit. We're not doing gross. It's not easy to calculate this, but you know, accountants can figure this stuff out. So, you get $200 in net profit. The net profit goes to the outside investors directly. The outi outside investors do not get shorted when a chip goes to Tesla and SpaceX. They get their full share of profit. Tesla and SpaceX do not get a share of the profit from those chips. If they sell a chip to someone else for the full thousand price, then the profit is shared evenly among all the parties. The reason for this is number one, you want Tesla and SpaceX to get priority. That's already in stone. But you don't want the joint venture to have some sort of incentive to try to undercut that and sell chips to someone else so they can get more money into the joint venture. By setting this up this way, the outside investors don't care who gets the chips. Tesla and SpaceX want the chips. So, if they want them, they get them. So, there's no real problem. Maybe there's an issue with management having the wrong incentives, but if Elon is the CEO of the joint venture, everybody's aligned. One of the goals when you do something like this is to have everyone's interest aligned. So this is a theory and again I want to stress this. This is my idea of how you can make a joint venture work that benefits Tesla, benefits SpaceX, benefits Alphaside investors and everyone's interests are aligned on here's what we're trying to do. We're trying to make as much chips as possible. We're trying to give SpaceX and Tesla an advantage in getting the chips that they need at a reasonable cost. And just to understand this, one of the huge advantages of Terapab is how much money it saves Tesla and SpaceX on chips. Right now, if you want to buy an H100, H2, whatever, GB300, whatever from Nvidia, you're talking about 30, 40, $50,000. The Vera Rubin, I think, is up around $50,000 for one chip. There's a whole thing that goes with it, but the chip itself is in the ballpark of $40, $50,000. I think Elon said that the AI6 was going to end up costing $6,000, but that may have been the market value. It's not clear whether he meant that would be the cost to manufacture the market value. I want to be clear on how they're saving money. When you make a chip, you have a chip design firm like Nvidia or Broadcom or AMD. They design the chip. They don't actually manufacture the chip. The FAB manufactures the chip. But Nvidia, AMD, Broadcom, they get a margin. They get a profit share from this. If it costs them $5,000 to make the chip and they charge $40,000, they're getting like a 70% margin or an 80% margin on that chip. So, not paying that 80% to the design firm is a huge win. Now, you get the fab. So, TSMC and Samsung are fabs. They own fabs and they make in the ballpark of 50 60% margin on making the chip. So, you start with a $50,000 chip and you shave it down to $5,000 if you take the design firm out and then you shave it down to $2,500 if you take the fabs margin out. So, you're probably talking about something like a,000 and $2,000, right? There's reasons why it's cheaper. It's half a die. There's there's a bunch of things, but if you So, you can see how you can get from $50,000 for a chip down to $2,500 for a chip, right? And that saves Tesla and SpaceX a crazy amount of money. And it makes it possible to sell the Cyber Cab for less money. If the Cyber Cab is even going to have these chips, CyberCAD might only have AI4. Optimus is running around with an AI6 chip. If the Optimus chip costs $50,000, you can't sell the robot for $20,000 or $30,000. But if the chip in Optimus only costs Tesla $2,500, now it's a lot easier to sell Optimus for $25,000. If SpaceX is going to deploy a AI satellite with a hundred chips and the chips cost $50,000 a piece, well, that's $5 million just for the chips on the satellite, not including all the other hardware that goes into it, the HBM memory and everything else. But if the D3 chips are probably $5,000 cuz it's a full dye, probably. If those D3 chips cost SpaceX $5,000 instead of $50,000, now you go from $5 million to $500,000 for the chips, you're making it so that you can put your AI satellites in space at very low cost and be able to deliver massive comput. And what this can also apply to terrestrial data centers, too, by the way. It saves money on terrestrial data centers as well. You're going to build out a terrestrial data center and you've got to buy $50,000 chips and you're going to put 300,000 chips in your data center, right? Because that's what they're doing. They're doing hundreds of thousands of chips. If you save $40,000 a chip on hundreds of thousands of chips, it's a lot of money. And it means you're now able to provide compute to customers at lower cost than anyone else because Google pays Broadcom and Google pays TSMC. Every one of these companies pays a chip design firm and they pay a fab margins. So if Tesla and SpaceX are able to get their chips without paying margins, they're able to build the lowest cost compute and serve customers in this tsunami of AI demand, this supersonic tsunami of AI demand. They're going to be able to serve customers at more scale, at lower cost than anybody else. And that wins the AI game. It wins the the real world AI, edge AI of humanoid robots and cars driving around and doing things in the real world. And it wins the data center cloud AI that many of us use in our daily lives. And you hit this wall where it costs too much and you've run out of time or it's not available right now because they don't have sufficient compute. Now all of a sudden it solves all these problems. And this is why it's not a merger, it's a joint venture. Terraab, Inc. probably not the logo. SpaceX and Tesla together partnering with outside investors. Maybe Intel, maybe it's the Saudis, doesn't really matter. People want to get in on the Elon train. They won't have a problem finding money for this. It's going to be gang busters. It's going to be huge. What do you think? Do you think this is a better idea than a merger? Don't you think they need that the value of generating cash so you can actually build the tariff ad is bigger than the value of a merger which generates no ga no cash and gives Elon control but doesn't give him the tariff ad that he wants. Loosely related to what we just talked about, I have a Tesla robo taxi and Optimus update indications that things are moving forward and we may see some developments with robo taxi soon. Somebody pointed out, and I just checked this myself on the Tesla website, that there are jobs posted for robo taxi fleet support. Robo taxi fleet support specialist. Palo Alto, they're already operating in the in the Bay Area. Tampa, Florida, they haven't been operating in Tampa yet. San Antonio, Texas, not operating in San Antonio, Texas yet. Orlando, Florida, not yet. Miami, Florida, not yet. So, this is coming. They're employing staff for fleet support. Robo taxi fleet support just to see how many jobs are related to robo taxi. You can see there was there's something like 30 plus jobs. I just decided to record scrolling down. Real estate manager for robo taxi architecture and design manager robo taxi supervisor fleet support operations manager. These are all over the country. All kinds of different lot of them in Palo Alto but they're all over the all over the place. San Antonio, Texas, Orlando, etc. many many different jobs for robo taxi support, robo taxi work. They are scaling up hiring people to make robo taxi happen. It's coming really soon. And there's more. This is just an example of what a job looks like. Highintensity actionoriented leader to join a robo taxi fleet team. A robo taxi fleet in Orlando, Florida. Ensure that our fleet of vehicles are properly maintained and charged while meeting the daily deadlines. a variety of what you will do in this job. And if that's not enough for you, there's 197 Optimus jobs. So, Robo Taxi is coming. Optimus is coming. I still don't expect to see meaningful financial numbers from Robo Taxi until the Q4 quarterly earnings call at the earliest, which is in January of 2027. Optimists probably pushed out later than that. So, I don't expect this stuff to jump to stock. But if you are holding out hope that your stock is going to go up, here's some hope. It looks like there's promising. This is real world AI that's going to make life better for everyone and especially for Tesla shareholders, of course, but for everyone. This is coming soon. Just a quick update on our AI stocks report. The AI stocks report, our primary index was up almost 1%. ASML, which is the lithography company that supplies fabs like TSMC and Samsung. They were up about 5%, Broadcom up about 5%. TSMC, AMD, a pretty good day for chip makers, except of course for some reason Nvidia. Palunteer down 5%, Tesla up a little. The big news though in our secondary index, Marll Networks up 32% in one day. As I understand it, Jensen Hong was on stage at Computex and said that Marll, which was about a $200 billion company, said they will be a trillion dollar company. So, the stock went up a third. Jensen Huang can move stocks. I can't move stocks. I don't know if Elon can move stocks, but Jensen Hang can move stocks. So, that drove everything up. Microsoft was down a bit. In our pod watch segment, one big pod to check out. Ethan Haye or Ethan He from XAI about XAI building Grock imagine video generation word models and video agents. Very excited. I haven't watched it yet. This is high on my list. Highly recommend it. I love this guy. I follow him on X. He's fantastic. Check that one out. And please consider supporting me as a YouTube channel member, an ex-subscriber, or a Patreon supporter. You get access to my options specials. I have a metric that I use to evaluate long-term call options and that helps evaluate not just long-term call options, but it helps you think about long-term investing. And there's some really good insights about these AI stocks that you can get from thinking about this that we talk about and we also talk about different strategies that might apply to these long-term call options. So, please consider supporting me as a YouTube channel member, as an ex-subscriber, or as a Patreon supporter. Thanks so much for watching.