SpaceX Has Real Value—But These 3 Stocks Have Better Odds Right Now

Key Points

• Altimetry Research values SpaceX at $1.3 to $1.5 trillion, suggesting roughly equal upside and downside at its current market cap.
• Analysts identify ASML, Northrop Grumman, and GE Vernova as cleaner alternatives to SpaceX for investors seeking AI and space infrastructure exposure.
• Tesla and AST SpaceMobile are flagged as stocks to avoid, with both trading at valuations that imply unrealistic future returns on assets.

The SpaceX IPO has investors asking a reasonable question: at a valuation somewhere between $1 trillion and $2 trillion, is the hype real, or is the market getting ahead of itself? Rob Spivey and Joel Litman of Altimetry Research say the answer is both—and that for most investors, there are cleaner opportunities sitting right next to the SpaceX story.

Altimetry applies uniform accounting to strip out the distortions in standard GAAP reporting. When Spivey and Litman ran that lens over SpaceX NASDAQ: SPCX's S-1, what they found was a profitable business with three distinct segments, each worth valuing on its own terms.

Breaking Down the Valuation

The launch business, what SpaceX calls its space segment, generates roughly a 12% return on assets, which is the U.S. corporate average, but strong for a capital-heavy operation with an 80%+ share of U.S. commercial launches.

That segment may double or triple in scale; Spivey estimates it could be worth around $120 billion.

Starlink is where the numbers get compelling. With $11 billion in revenue and a 30% return on assets, three times the corporate average, it already looks like a premium business. The addressable market isn't the entire $1.5 trillion global telecom industry, as Elon Musk has suggested, but a more realistic $100 billion opportunity still implies a 10x growth runway. Comparable satellite businesses have reached 60% returns on assets. If Starlink gets there, Spivey puts the segment at around $600 billion.

The xAI compute segment is where the bull case hinges. It's losing money now, but recent agreements to supply compute capacity to major hyperscalers suggest a path to becoming the go-to infrastructure partner for AI scaling demand. In a strong-case scenario, that segment could be worth $600–$700 billion on its own.

Add it up, and Spivey sees a range of $1.3 to $1.5 trillion—not the irrational number skeptics claim, but not the discount that buyers would want either. "I could see as much downside as upside from here," Spivey said. His position: don't short it, but don't chase it.

The Irreplaceable Equipment Maker

For a better entry point in the AI infrastructure story, Spivey points to ASML Holding N.V. NASDAQ: ASML. The Dutch semiconductor equipment maker produces the extreme ultraviolet lithography machines that Taiwan Semiconductor NYSE: TSM and other chipmakers need to manufacture the most advanced chips, for NVIDIA NASDAQ: NVDA, memory providers, and anyone else building at the frontier.

The moat is genuinely hard to overstate. ASML is the only company in the world that builds this equipment, and the machines are so complex that a Chinese manufacturer that bought one, disassembled it trying to reverse-engineer it, and ultimately had to call ASML back to put it together again. That story tells you something real about competitive position.

On a uniform accounting basis, ASML's return on assets runs closer to 22%—not the 15% that as-reported figures suggest. The company is in the middle of a major production ramp, shifting its mix toward higher-end equipment that carries fatter margins. That mix shift is why it has consistently beaten earnings expectations. And as AI infrastructure build-out continues, whether data centers are terrestrial or eventually orbital, demand for ASML's equipment only grows.

The stock trades at a high nominal price, but Spivey's argument is straightforward: price per share is irrelevant. What matters is the earnings relative to that price, and on a uniform basis, ASML's forward P/E is closer to 45x against 20%-plus earnings growth potential. For investors who can't get comfortable with the share price, fractional shares are the practical answer.

The Space Defense Sleeper

The second buy on the list looks nothing like an AI play at first glance.

Northrop Grumman Corporation NYSE: NOC is a defense giant, but a quarter of its business and a disproportionate share of its profits come from space.

The company builds satellites and has a launch capability that most investors don't associate with the brand.

Spivey's case is that SpaceX's success isn't just good for SpaceX. It's catalyzing the entire space economy, pulling more commercial and government investment into the sector.

Northrop Grumman is one of the few contractors with the technical depth and established relationships to capture that demand.

It also benefits from the broader defense spending story—U.S. budget increases, allied spending ramp-ups, and the need to replenish equipment used in recent Middle East operations. The administration's stated priorities—blue-water Navy, missiles, space—map directly to Northrop Grumman's core business.

As-reported numbers make Northrop Grumman look like a barely-above-cost-of-capital business at 5–6% return on assets. Uniform accounting puts it closer to 12%. With a uniform P/E of around 26x and the market pricing in roughly 5% annual earnings growth, Litman and Spivey see that estimate as far too conservative given the tailwinds. The stock doesn't need a massive multiple expansion; earnings growth alone could drive meaningful upside.

The Power Bottleneck

GE Vernova Inc. NYSE: GEV rounds out the buy list, and this one is directly tied to the AI power race. Data centers can be built in 15 months. The large gas turbines—the heavy industrial equipment that generates the electricity those data centers need—have a five-year-plus backlog. GE Vernova makes them.

That supply constraint gives GE Vernova pricing power that has fundamentally changed its economics. Historically, the company operated on a razor-and-blades model, selling turbines at thin margins and capturing value through decades of service contracts.

Now, with demand far outstripping supply, it's getting premium pricing on the hardware itself. Margins are expanding from what was a 5% return on assets when the company spun out of GE toward the 20%+ range that industry peers have achieved.

The company is simultaneously ramping turbine capacity—targeting 20 gigawatts in 2024 compared to 10–12 gigawatts before the spinoff, potentially reaching 30 gigawatts—and doubling capacity in its electrification segment, which covers the transformers and grid infrastructure needed to move power from generation to the data center. The recent earnings beat was driven by exactly this combination: higher volumes and better margins landing together. Neither trend is short-term.

2 Names to Avoid

AST SpaceMobile, Inc. NASDAQ: ASTS is the name that investors in the direct-to-cellular satellite space have been watching closely. The company's ambition, enabling satellite connectivity through standard mobile phones, without specialized hardware, is compelling.

Its challenge is that Starlink is already pursuing the same market from a position of profitability and a denser satellite constellation, while AST SpaceMobile is currently burning cash at a negative 14% return on assets.

The bigger problem is valuation. At roughly $30 billion, the market is already pricing AST SpaceMobile to achieve 20–30% returns on assets, which essentially assumes the company captures a dominant share of the global direct-to-cellular market and that carriers like Verizon Communications Inc. NYSE: VZ and AT&T Inc. NYSE: T back it exclusively over Starlink. Spivey's view is that AT&T and Verizon will want AST SpaceMobile as a competitive check on SpaceX—a stocking horse, like Advanced Micro Devices NASDAQ: AMD relative to Intel NASDAQ: INTC—but that role implies a ceiling on the business, not a path to $100 billion-plus.

Tesla, Inc. NASDAQ: TSLA is the harder case to make, given how many investors have deep conviction in Elon Musk's long-term vision. Litman's argument isn't about the vision. It's about what the stock is already priced to deliver. To justify the current valuation, Tesla needs to achieve a 50% return on assets—higher than Coca-Cola NYSE: KO, higher than almost any company in any sector, in a business where Chinese manufacturers compete with government-backed subsidies, near-zero IP enforcement costs, and pricing that Western automakers can't match.

Tesla's current return on assets is around 6%, which is reasonable given the competitive dynamics. Getting from 6% to 50% would require a near-complete elimination of that competition or a business transformation—through Optimus robotics or other ventures—that neither history nor current economics supports. Even in the bull case, Litman sees Tesla as a good industrial business, worth a 20% return on assets. That's still far from what the market is pricing.

The SpaceX merger scenario, in which Musk rolls Tesla's battery and robotics assets into the broader SpaceX ecosystem, could change the math. Spivey and Litman say they wouldn't be surprised if it happens. But that's a different bet than the one Tesla shareholders are currently making.