There was a moment, sometime around 2015, when Elon Musk was the most genuinely exciting person on the planet. Not exciting in the way celebrities are exciting, like, it wasn’t about who he was dating or his slate of upcoming films, he was exciting in the way that made you feel the future was going to be OK.
He was building reusable rockets. Making electric cars cool. Talking about Mars with the straight-faced conviction of a man who had made the maths work. Whether you believed him or not, the ambition was intoxicating. He was, as far as many of us were concerned, the closest thing the real world had to Tony Stark.
I’m writing this the day before June 12, 2026, when SpaceX goes public on the Nasdaq at US$135 (RM548) a share. The valuation is US$1.77 trillion (RMRM7.19 trillion), which would make it one of the most valuable companies ever to go public in human history. Bigger than Tesla. Bigger than Saudi Aramco at its IPO. But is SpaceX worth the market cap?
There is a real business here. SpaceX controls over 80% of US rocket launches. Starlink, its satellite Internet service, has 12 million subscribers across 160 countries and is actually profitable. The engineering achievements are among the great industrial accomplishments of the last 20 years. Nobody is disputing any of that.
What they are disputing is the number. To justify US$1.77 trillion, SpaceX has to grow at a rate no company in history has ever sustained. It posted US$18.7bil (RM76.1bil) in revenue last year but also a US$4.9bil (RM19.94bil) net loss. Its artificial intelligence (AI) division burned through US$7.72bil (RM31.41bil) in the first three months of this year alone.
The company claims a total addressable market of US$28.5 trillion (RM115.98 trillion), a number so large it essentially means: trust us, the upside is everywhere. When you buy the stock, you are not buying what SpaceX is. You are buying what Musk says it will become.
Part of what he is selling is the orbital data centre dream. Not server farms in Johor or Singapore but a constellation of up to a million satellites in low Earth orbit, processing AI workloads in space, powered by near-constant solar energy. The cooling, the solar power, it’s hard on Earth but it’s easy in space.
There is a slight issue: Almost none of the infrastructure required to do all that exists yet. The dream is being sold before the plumbing has been invented.
This is not the same as building an electric car in 2010. When Tesla launched the Roadster, the technology required was hard but known – batteries, motors, software. You could hold the thing and drive it. Building data centres in orbit requires solving entirely new categories of engineering problems: chip hardening for radiation, thermal management in vacuum, satellite-to-ground data transfer at scale.
SpaceX has done things people said were impossible. But data centres in space seem like something that is further away than Musk likes to acknowledge, and that matters when you go public with a company worth US$1.77 trillion. But Musk is nothing if not optimistic.
He recently said that he is building an AI chip two to three times better than the famous Nvidia one at 10% of the cost. Nvidia commands a US$5 trillion (RM20.35 trillion) market cap and posts over US$80bil (RM325.56bil) in quarterly revenue because they make the best chips on the planet.
Musk said he can “visualise” the new chip. Not that it’s been visualised but that he can visualise it. Will visualising things keeps investors happy with SpaceX?
Which brings me back to this point: Elon Musk of 2026 is not the same man who made you believe in electric cars and reusable rockets. The last two years have seen him run a “Doge” to make the US government more efficient with very questionable results. He has become one of the most polarising political figures in multiple countries, and presides over a Tesla brand so damaged in some markets that owners are putting disclaimers on their cars. And he’s been promising full self-driving cars since 2015 but they have yet to arrive.
I will watch the opening bell tomorrow with something between excitement and unease.
The rockets are real. The engineering is extraordinary. The dream – Mars, orbital computing, humanity as a multi-planetary species – is one of the great dreams on offer. But at US$1.77 trillion you are not just buying the rockets. You are buying a chip that does not exist yet. You are buying a million data centres in space with no engineering to put them there. And a version of Elon Musk that may or may not look the same in five years.
Will SpaceX live up to the dream for investors? It’s a cliché but only time will tell.
