Gene Munster Says SpaceX Is the Only Sovereign AI Company and Should Be a "Core Tech Holding." Time to Buy While It's Below $150?

Two months before the massive $86 billion raise from the Space Exploration Technologies (NASDAQ: SPCX) IPO, former Apple analyst and co-founder of Deepwater Asset Management Gene Munster said in an investor note that SpaceX was "the only entity in the world building sovereign AI."

That phrasing does not imply that Munster thinks SpaceX's AI will rule everything, but rather that the company is building and owning all the artificial intelligence software and hardware it would require in such a way that essentially no other company could bottleneck its progress.

Munster is bullish on the company for this reason, and he said recently that SpaceX should be a "core tech holding" for investors. So, should investors follow his opinion and buy SpaceX now?

Image source: Getty Images.

SpaceX is doing something not even Google has pulled off

Munster's definition of sovereign AI in SpaceX's case is an infrastructure consisting of its rocket launch capabilities, its Grok AI model, its Starlink satellite broadband service, and its planned Terafab semiconductor foundry. That combination of assets gives SpaceX an AI edge that not even Alphabet can match (since the Google parent can't launch its own orbital data centers into space).

As Munster said in the investor note, Alphabet must rely on the chip manufacturing capabilities of Broadcom and Taiwan Semiconductor (also known as TSMC), and it doesn't have an in-house rocket launch operation:

Google still relies on external fabrication (Broadcom/TSMC), doesn't own last-mile network delivery, and lacks a launch vehicle to deploy infrastructure off-world.

SpaceX's potential advantage over its AI competitors hinges on a couple of important things, though. First, it will have to achieve its goal of drastically reducing launch costs with its Starship rocket compared to its Falcon rockets. Second, it will have to work out all the complexities of getting a constellation of data center satellites operational in a commercially viable way.

No company has orbital data centers right now, and even with the advantages it has from being able to launch its own rockets to deploy such satellites, there's no guarantee SpaceX will be able to get orbital data centers to work or that they'll be cheap enough to be useful.

What's more, SpaceX has barely broken ground on the Terafab site where it aims to build the semiconductors for its data centers. The idea is that once that foundry is churning out silicon at scale, SpaceX won't have to rely on leading chipmakers like Taiwan Semiconductor and Broadcom, which have many large customers, and whose production can get bottlenecked when demand is high (as it is now).

The Terafab project will include most of what SpaceX will need for its data centers, including chip design, wafer fabrication, and even memory processors. That latter category of chips is important, as AI software has massive memory demands, and a shortage of memory chips has led to soaring prices.

But SpaceX will still rely on other companies during the Terrafab setup, most notably Intel, which is helping it set up and build its fabrication infrastructure. And even if the Terafab project works well, there's no guarantee that its orbital data center plans will succeed.

For reference, analysts at Morningstar say SpaceX couldn't launch commercially scalable orbital data centers until 2028 at the very earliest, "even in the most optimistic scenario."

Still, Munster argues that if SpaceX can get them to work and they're efficient, no other company will have the AI infrastructure advantage that it will have.

Is it time to make SpaceX a core tech holding?

Even with all of SpaceX's potential, I think it's too early to buy the stock. The biggest reason for this is that SpaceX just went public and will likely remain very volatile for at least the next year.

Jeffries' research over the past 20 years shows that companies that go public with market caps of $10 billion or more average returns of only 3.5% one year after their IPOs.

What's more, SpaceX shares are trading at a hefty premium, with a price-to-sales ratio of 110. The tech sector average is just 9.

Meanwhile, the company is ramping up its spending -- capital expenditures were $10 billion in the first quarter, nearly a third of what SpaceX spent last year. And it incurred nearly $5 billion in losses last year.

The point is that SpaceX stock is expensive, pursuing the company's ambitious goals will be extremely costly, and most mega-IPOs prove disappointing in their first years on the market.

Even if Munster's bull case does eventually pan out, waiting on the sidelines is likely the best move for now. I recommend watching the company over the next year or so to see how well it progresses toward its goals before making a decision on whether or not to buy SpaceX stock.

Chris Neiger has positions in Apple. The Motley Fool has positions in and recommends Alphabet, Apple, Broadcom, Intel, Jefferies Financial Group, and Taiwan Semiconductor Manufacturing. The Motley Fool has a disclosure policy.