$TSLA · InvestAnswers · 2026-11-05

today we're going to talk about uh Tesla, SpaceX,

We talk a lot about Tesla today as well. And both are down hard.

Tesla looks like it bottomed around 345. It's moving back up a little bit slowly right now. It's not exactly a V-shaped recovery.

imagine Tesla cars all of a sudden do a lot of inference and they can run a lot of compute very cheaply.

Looks like maybe double bottom could be in for Tesla SpaceX. people think Tesla and SpaceX are doing bad. You should look at some of the other names out there getting absolutely smashed.

When do Tesla cyber cabs hit 10,000? Now, my my tipping point was always 2,000 cyber cabs because then Wall Street will wake up and see even 2,000 cyber cabs running 8 to 12 hours a day generating$1 to2 dollars a mile moves the needle,

So we're already at about 1,400 um unsupervised cyber cabs in Texas and Florida. So they have registered like I haven't checked today but 1244 as of yesterday. Uh and then I believe that there's um about 150 to 200 in Florida because um Florida miles equal Houston miles.

The tracker is not showing those because there's you know like no tracking in Miami. Uh, so they're doing as many miles or more miles than Houston. So Houston is showing 70 vehicles there or like 47 or something like that.

So there's um they don't that that tracker only does 1% or less of the miles. So the increase in miles indicates so they're already at 1400. I expect them to get to 2,000 this month um because they'll put in 300 cyber cabs into Houston and Dallas.

So that' be another 600 year. 2,000 right there. So 2,000 this month I think quite for sure. and then um 10,000 you're already at um um like a thousand in Austin and that's in an area that's still onetenth of the metro area.

You haven't even got to you're still 20% away from Austin um the main core uh Austin, not the metro area. So you could put in 4,000 vehicles into the full Austin metro area and then we've increased the vehicles to 200.

So you can other other five other six cities you could then get to 40,000 vehicles to fully fill it up. Even just partially filling it up gets you to 10,000 I think in January.

I don't think you would I wouldn't say you destroy them. I still think there would be room for them at 4,000. At Definitely 2,000 there's still plenty of room for them. Um I think 4,000 is still fine because the level you need to pull totally um dominate and and to undercut them and push them out would be probably about um 10,000 into the metro uh Austin area.

Um, but if you're saying you want to preserve that, you want to like not get to that fight now, then you stay about like 2,000 and definitely 1,000. They're they're roughly at that now already.

And um, way more Austin and the ride share fine.

Cyber Cap has been spotted in 20 different states. and Tesla has been proven to be a lot safer than Whimo. So, we we'll see where that goes. But I I personally I've gotten to the stage now uh like 99% of my last 10,000 miles in my Cyber Truck, all FSD.

My insurance record 100% perfect, all safety because of that. Because you're now incentivized not to drive because your insurance goes down. They really they really have thought of everything. So that's good news.

They are testing the semis for their version of FSD. I think that would save so many lives because I I drive over a mountain pass on 80 and I see a lot of trucks sleep at the wheel, veering lanes, losing control in ice. We're in the icy season now. very dangerous.

and I do know as well that the regenerative braking is much better to handle in the ice as well than physical brakes. Um so 10 million robbo taxi that would be um late 2028 or 2029.

Um I think you know ideally perfect execution super roll out unrestricted um there's 4 million or 5 million um per customerowned vehicles that could be put into the fleet. So then if that became a money-making thing even if some people didn't want it if they sold to other vehicles.

So doing all those vehicles and then adding another 1 million cyber cabs, you can get to like 5 million. So if you add 2 million cyber cabs in 2028 plus um another million um Model Y's and then more of the the people join it, then you get to the 10 million as early as 2028 in a perfect scenario.

Um for I think we still get to like I'm pretty confident get to a million next year um end of next year.

So I think robot trucking is better than robo taxi in terms of like actual money because all semi-truck miles all truck miles are paid everywhere around the world. It's about two bucks per mile.

There's there is no like I have my own personal truck for something else. It's like they're all commercial and half of the cost of trucking is the driver,

44% to 50% depending on the country you're in. That's right. That's right. So half the cost is the driver. Um, so then if you convert all 40 million vehicles to that and you'd want to and the robot trucking is 10 times cheaper than regular trucking and it would also increase the volume and it it is strategically so one it's huge commercially and you can make a half million dollars per truck.

So if my truck cost in a year I can make a half million. I can make 10 times more money than one robo taxi with one truck. So the economics are just uh wonderful on that for it.

And when we do robo trucking, people talk about affordability. The cost of goods, if you're not paying Amazon Prime, if you're paying UPS to move your stuff around or other things, it's about 10% of the of your product cost goes into transportation.

So if I make it 10 times cheaper, I'm giving like a 9% value added tax cut to everything around the world. So if they have the adoption in United States, in Europe, uh in most of the trucks are in Asia 80% because they've been the factory to the world.

Although if you're onshore that would change things, but it's just a huge opportunity.

right? I'm assuming that all of the um people who already have their vehicles will then reconsider and get FSD or they will sell their vehicles to someone else who wants FSD because the only path to FSD is with the Tesla hardware for vehicle.

The latest version 15 is a hardware 4 vehicle. So, I'm expecting the adoption rate to we're already at uh 75% adoption rate um on new vehicles. I expect that to go towards like 90 or 100% on new vehicles.

And then on the old vehicles were at 30% adoption rate. I think that goes up to you know the 80 90% level in US market could be slow in Europe. China would would then go there as well.

Yeah. And the math is very simple. If you take 6 million times 100 bucks a month, 95% margin. It's extremely lucrative business. And then on top of that, you got the robo taxi revenue.

Um remember each car technically in the US in an expensive taxi area which is where they're deploying um each car can make 50,000 a year 1 million robo taxis takes us to 50 billion.

But what people what people get tripped up on and this is the math they see ah 50,000 per car yeah cars cost a lot but no when robot cyber cab gets to scale they can pop them out at 15k a pop.

So you're generating 50,000 per year on a one-time 15k cost. Of course, the service cleaning and etc. on top of that, but I'm sure they do that at scale, too.

This is where things get really wonky. And what I'm always trying to calculate is where do you play? Which which one will make more infinite ROIs as I call them faster? Will it be SpaceX with compute or Tesla with FSD and robo taxi?

So the $50 billion here for 1 million robo taxi that's the revenue you get from gunrange and spam. Gun range and spam B300s you get $50 billion from those two uh data centers. Um if you're saving some for training maybe three include mini hard into that.

So the the the data cent we built in one quarter complete in one quarter um is equal to this revenue. Um but you know this revenue is still great but it is not as good as um the data centers.

what about variable costs because that's the big factor. So when I take imagine you take okay you're making a 50 billion from 244 440 megawatt facilities but your chips are advertising at 20% per year and there's $30 billion of chips maybe in those two facilities.

Is that approximately correct? Sounds about right. Yeah.

Yeah. Uh so that's six billion a year in amortization taken off of that. You got the variable cost runs about 20% between electricity and cooling and the team on the floor. So that eats into that.

But with the robo taxi revenue, it's the margins appear to be higher as far as I'm concerned from the actual upfront cost.

Yes. Yes. You're true. that that that would be a different um we have to do a more detailed analysis to to see what it is. Um but still the the overwhelming speed of the AI data centers and getting if I get to 10 gigawatts would overwhelm this number, you know, even with only 50% margin, you would overwhelm this number. Could you go to 500 billion?

Optimus, real volume is still demos. We know Fremont is built for 20,000 a month if I'm not mistaken. That's what's that 200,000. That's a million. A million a year. A million.

Yeah. And then we have other things like FSD just hit 1.7 million paid FSD. A little above 1.7 million paid FSD users. Your target for 2027 is 6 million. That's more than tripling. That's very aggressive,

Yeah. Well, they're they're built for a million a year, but the initial scale is 20,000 a month. It's going to take time to get there. That's not going to be fast. That's be next year supposedly for by time.

So they might be pumping out 200,000 of them by the end of next year. That's my guess probably.

The first first question is where do they go? He did talk about 20,000 of them going into the academy to learn from each other like like a play school kindergarten type of thing.

Um the some will go to the Tesla factories and then the next stage will be Tesla suppliers I reckon.

I don't I don't see domestic use coming for a while. Walk us through your theory. So my theory about the home is it's all about the the the chip for the AI because I have an AI5 chip in there that is like a pedawatt of compute.

So that's like um 20 times or 50 times the compute of a laptop or of a small server. So people if I want my local AI the best option you know would be to get an an Optimus for that.

Even if it just like stood there and did nothing it in terms of like physical one that uses hands, you know, just using this camera, something like that. The fact that it could do um uh compute means I could do edge computing, I could do that.

So the edge computing delivery for an Optimus that also can walk around, talk and have all these other sensors is makes it the best edge computing uh device. So, so I'm skipping o away from the um the the usage of it for the hands like that use case is unnecessary to sell the units because you would buy it for that that purpose but you know I think the utility of it to use hands is a cherry on top that makes um penetration easier.

So it's it's less a concern about the hands but um so I think maybe it be the people who are buying um you know developers or companies could buy it and then put it into white collar work.

but yeah warehouses easy thing um factories space will do itself but then um some level home usage just for the chip is my use case value argument. Yeah. And then V3. I know two people that have seen V3, but they're sworn to secrecy.

It's out there. We don't know when the demo is going to happen, but it sounds like if they're pumping out 20,000 of them a month in the very near future, it's going to be hard to hide them. So,

I think I think the reveal is coming. And we know why they're afraid to reveal is because they don't want the Chinese copying them. That's the situation. Oh, one thing is we're short of compute for Grockbot.

So if you have a Grockbot on your your humanoid bot, then you have your digital agent working with your physical agent and you have that stream of stuff there. So it's um it's another up level of agents um with compute.

All righty. So, um, Optimus is coming. Um, when do you see that being a needle mover for the business? Um, it depends upon your theory. If if you do a great demo and have it to 10,000 units that are being deployed, that I think that should be a clear need mover.

Um yeah. Yeah. I think that that's the main level I would like to see.

Your forecast for a merger for the end of the year is 30%. Has that changed with the change in the political environment? Um yes, I think it it increases the chance slightly to try and announce, but it's it's also um bureaucratic complications.

So, it's like the urgency, as we discussed at the top, it's urgent that they do it, but it's very also very challenging and difficult. So, I I'm not in there knowing how complex and difficult it is.

I just know that they if they're going to do it at all, they need to do it relatively quickly. Um, I think by Q1 definitely. Um, so if I was to put a bet on it, I'd be saying January that they announced it.

That's I I actually believed it brings more urgency to kind of put Tesla under the umbrella of national security for the United States. Now uh people are still confused, tripped up by the deal.

Basically the quick summary is SpaceX already have issued the shares to acquire Tesla. That was all in the S1 we saw a long time ago. SpaceX will buy Tesla. There will be additional shares.

But people need to look at the synergy between both. Imagine they're both two trillion market cap. You get a combined entity with synergy of 4 trillion. The increased number of shares gets divvied up.

We were at a 3:1 ratio of SpaceX shares to Tesla just last week. Um, I haven't checked the chart, the pair chart of what it is right now today with all the moves, but let's look.

Actually, Tesla is down, they're both down exactly 10.7%. Tesla and SpaceX. So, they're moving in lock step, which I forecast a while back because of this acquisition.

Um, and then people are also concerned about what happens to SpaceX leaps or Tesla leaps. And the answer is they get rerated in the new version. But be careful if you are buying leaps.

Do not buy out of the money leaps. Make sure they are kind of in the money or else it can be much more risky.

Uh anything to add on the deal terms of this merger that you believe has a much more accelerated chance of happening.

imagine no no merger happens. Tesla standalone to 2030 mark cap is 10 trillion. That's based mo mostly off of the cyber cab ramp and valuation. We saw how lucrative that business is.

FSD uptake and a little bit of optimism, not too much optimist.

So you can look at the market caps and they're nearly in lock step from where they are today with the multiples. So Tesla no merger would be a 6.6x from here going to 2530 a share which is bonkers to think about just four years away.

I want to know what I should buy if I'm DCAing into Tesla or SpaceX. Which one is cheaper relatively now? I think I'd be very happy if the merger happened at a ratio of three. That means you get three shares of SpaceX for every Tesla share.

Tesla 348 it fell as low as 340.

How does Tesla and SpaceX get their hands on those supplies? Or are they creating a new type of solar panel? I think that's my theory, the latter. They're doing something different that doesn't rely on traditional supply chains.

No. No. because we didn't even talk about the fact that Terraab is a joint project between Tesla and SpaceX and would become equal to all of TSMC's uh chip production although they may TSMC may be a partner in it.