SpaceX fell 13.61% on Aug. 5, the day after the company’s first quarterly earnings report as a public company landed, unnerving investors with capital expenditures running more than $5 billion above Wall Street estimates.
The stock has now declined for five consecutive weeks. The conversation on Wall Street has shifted from when to buy to whether to buy at all.
Jim Cramer has a different question entirely. On the Aug. 5 episode of “Mad Money,” he asked investors to stop thinking about the next quarter and start thinking about the next century.
What Cramer actually said about SpaceX on “Mad Money“
“Back in the day, people bought 100-year railroad bonds that paid off,” Cramer told viewers, according to CNBC. “SpaceX could be a 100-year piece of paper too.”
He then asked the question that got the most attention. “Do you think your children or your grandchildren won’t be doing stuff on the moon someday? Do you think that orbital data centers won’t make sense? They can get all the power from the sun. Might end up being the only place politicians let us put them. Think about it. Maybe you put some away for the next generation, or even the one after that.”
More SpaceX:
- Morgan Stanley says SpaceX investors miss the bigger story
- Peter Schiff says SpaceX is a warning for hyped stocks
- SpaceX stock defies latest Wall Street forecasts
At the CNBC Investing Club’s morning meeting the same day, he went further. “I would give my kids this stock. This man is a visionary,” Cramer said, as CNBC reported.
The railroad bond comparison is deliberate. Those bonds funded infrastructure that took decades to pay off but ultimately became foundational to the American economy.
Cramer is making the same argument for SpaceX, that the company is building infrastructure for markets that barely exist today but could define the economy of the next generation.
Why Cramer said to wait before buying SpaceX stock
Cramer’s long-term enthusiasm didn’t translate into a buy-now call. He had been warning investors about this exact moment for weeks. In late July, before the earnings report and before the lockup expired, he told viewers to hold off on building a large position.
“Generally speaking, more supply results in lower prices,” he said, pointing to the 911.5 million shares that would become eligible for sale when the lockup ended on Aug. 6.
That expiration effectively doubled SpaceX’s public float, from roughly 639 million shares to 1.55 billion, or from about 5% of shares outstanding to just under 12%. At current prices, more than $100 billion in stock became available to trade for the first time in a single day, according to Reuters via Yahoo Finance.
Cramer’s message is that the long-term thesis and the short-term entry point are two separate conversations. Believing SpaceX is a generational investment doesn’t mean paying any price at any time to own it.
SpaceX Starlink, AI compute, and orbital data centers, explained
Cramer’s thesis rests on SpaceX developing multiple businesses, each of which could become significant on its own. Here is what he is actually betting on.
Starlink, AI leasing, Starship and orbital data centers in Cramer’s case:
- Starlink. Already the company’s most profitable segment, with 12 million subscribers across more than 160 markets. Starlink posted a $1.66 billion operating profit in Q2 2026. It’s the current engine funding everything else.
- AI compute leasing. SpaceX is the only profitable segment from Starlink, with 12 million subscribers. The AI segment generated $2.56 billion in revenue, up 247% year over year. Anthropic is paying $1.25 billion per month for the Colossus 1 data center. Google is paying $920 million per month for GPU access. Management said $6.7 billion in new cloud deals were signed in early Q3 alone, over a six-month period beginning in October.
- Starship and launch services. The reusable rocket system has the potential to dramatically reduce the cost of sending cargo into orbit. SpaceX handles more than 80% of all mass to orbit since 2023. Every cost reduction strengthens the economics of everything else the company does.
- Orbital data centers. No company has proven this works at scale. But Cramer’s point is that the economics could eventually favor space-based computing, with constant solar power, no local land or water restrictions, and no political opposition. This is the most speculative part of the thesis and the one furthest from generating revenue.
- Lunar commerce. Also speculative, but SpaceX is the primary contractor for NASA’s Artemis lunar lander program. If humans return to the Moon and establish a presence there, SpaceX is positioned to be the infrastructure provider for whatever commerce follows.
Justin/Getty Images
George Noble’s $30 SpaceX target and the bear case against Cramer
Not everyone on Wall Street is thinking in centuries. George Noble, a veteran investor and former Peter Lynch associate, holds a $30 fair-value estimate on SpaceX, implying 72% downside from current trading levels.
His argument is that the “Elon Musk premium” embedded in the stock’s valuation could become a discount if the capital-intensive strategy falls short of its ambitions, according to Stocktwits.
SpaceX’s Q2 capex hit $18.4 billion with $15.83 billion going specifically to AI infrastructure. Management guided for similar spending levels in Q3 and Q4, putting full-year capex on track for roughly $65 billion against a Wall Street consensus of $50 billion.
Free cash flow remains negative. The AI segment posted an operating loss of roughly $1.3 billion, even as AI revenue grew 247% year over year.
The railroad analogy only holds if SpaceX eventually builds businesses that actually pay. Not every 100-year bond paid off. Some railroads went bankrupt. Starlink could plateau. AI leasing could disappoint. Orbital infrastructure could turn out to be economically unviable for decades longer than anyone expects.
The analogy is useful, but it isn’t a guarantee.
What Cramer’s SpaceX railroad bond call means for SPCX investors
Cramer is not telling investors to buy SpaceX today at any price. He is telling them to think about the stock differently than they would think about a company reporting next quarter.
The railroad bond framing is a way of saying: If you are going to own this, own it the way people owned infrastructure bonds in the 19th century, with a time horizon measured in decades and an expectation of significant volatility along the way.
For investors who accept that framing, the current weakness, the lockup selling pressure, and the capex concerns could look different. Not necessarily a reason to buy immediately, but potentially a reason not to dismiss the stock entirely because of what one earnings report said about one quarter.
That question doesn’t get answered this quarter. It probably doesn’t get answered next year, either.
SpaceX either builds something that justifies the spending, or it doesn’t. Cramer thinks it will. George Noble thinks it won’t. The next generation gets to find out who was right.
