Two trillion dollar SpaceX IPO Mars bet

**SPEAKER_1** (0:00)
SpaceX is targeting a fixed price of $135 per share for its initial public offering. They are seeking an overall valuation of roughly $1.75 trillion to $2 trillion, and they're aiming to raise up to $80 billion.

**SPEAKER_2** (0:17)
An $80 billion capital raise completely dwarfs the previous global record. I mean, Saudi Aramco held that. But usually an equity event of this size relies entirely on sovereign wealth funds or the largest institutional asset managers in the world.

**SPEAKER_1** (0:33)
Right. And the strange part of this specific structure is that retail investors are suddenly being offered 30% of the entire allocation.

**SPEAKER_2** (0:42)
Which is highly unusual for something this large.

**SPEAKER_1** (0:44)
Yeah, extremely. So with a fixed price tag, highly complex allocation rules across multiple consumer brokerages, and a corporate structure intertwined with an artificial intelligence and space infrastructure, the real question is whether everyday investors will actually secure a piece of this company at that $135 price, or if they're walking into a highly orchestrated financial trap.

**SPEAKER_2** (1:05)
Well, the financial reality inside the regulatory filing is really the only place to start looking for that answer. Because the top line numbers show two very different companies operating under a single roof.

**SPEAKER_1** (1:16)
Yeah, they are not currently profitable on a net basis. You look at a trailing net loss of $4.9 billion against $18.7 billion in revenue.

**SPEAKER_2** (1:27)
Right, so you have Starlink generating the profit, and then you have Starship and the artificial intelligence infrastructure basically consuming it.

**SPEAKER_1** (1:34)
Starlink is this highly predictable recurring revenue stream. I mean, they have over 10 million active subscribers paying a monthly fee for satellite internet. That generated over a billion dollars in operating profit for a single quarter.

**SPEAKER_2** (1:49)
It looks exactly like a standard, high margin telecommunications business.

**SPEAKER_1** (1:53)
Exactly, but the capital expenditure side of the ledger is where that traditional telecom comparison completely ends.

**SPEAKER_2** (1:59)
Because building a fully reusable interplanetary transport vehicle like Starship requires continuous iterations. You have explosive testing, entirely new manufacturing supply chains just for the rocket engines.

**SPEAKER_1** (2:10)
Yeah, that process burns cash at an industrial rate. And then you add the $12.7 billion expenditure on artificial intelligence infrastructure following the merger with XAI.

**SPEAKER_2** (2:23)
If you are wondering why a space logistics company is spending $12 billion on artificial intelligence data centers, you really have to look at the engineering bottlenecks they face.

**SPEAKER_1** (2:32)
Right, because modern rocketry is essentially a physics simulation problem at this point.

**SPEAKER_2** (2:37)
Yeah, designing an engine bell or modeling thermal dynamics during atmospheric reentry, that requires running the Navier-Stokes equations for fluid dynamics over and over again.

**SPEAKER_1** (2:48)
And doing that with traditional computing takes months.

**SPEAKER_2** (2:51)
So by building a massive internal computing cluster with XAI, they can run those physical simulations using machine learning models in a fraction of the time.

**SPEAKER_1** (2:58)
And that same computing power is also necessary for Starlink.

**SPEAKER_2** (3:01)
Right, because when you have thousands of satellites in low Earth orbit communicating with each other using optical lasers, the routing protocol is not static.

**SPEAKER_1** (3:10)
No, it's constantly changing. The network has to calculate the most efficient path for a packet of data to travel from, say, London to Tokyo. It bounces across a moving grid of satellites while factoring in orbital decay, atmospheric interference, which requires localized artificial intelligence just to manage the traffic. But the market is being asked to apply a hundred times revenue multiple to this combined entity.
That kind of multiple is typically reserved for early stage software companies with near zero marginal costs.

**SPEAKER_2** (3:43)
Right, you write a piece of enterprise software once, and you can sell a million licenses with almost no additional manufacturing expense.

**SPEAKER_1** (3:51)
Applying a software multiple to a heavy industrial manufacturing business requires a total suspension of traditional valuation models.

**SPEAKER_2** (3:59)
Rockets and satellites have enormous marginal costs. I mean, they operate in a physical environment where hardware occasionally explodes on the pad.

**SPEAKER_1** (4:06)
And where thousands of low earth orbit satellites need to be physically manufactured, launched and then continuously replaced as gravity pulls them back into the atmosphere to burn up.

**SPEAKER_2** (4:15)
You are basically paying a cloud computing premium for a very harsh industrial reality.

**SPEAKER_1** (4:21)
The internal logic of that $2 trillion valuation seems directly tied to the compensation structure outlined in the filing.

**SPEAKER_2** (4:28)
Right. The founder receives 1 billion Class B shares. But those shares only vest based on specific market capitalization milestones.

**SPEAKER_1** (4:37)
And one of those explicit milestones for unlocking the shares is the establishment of a permanent human colony on Mars.