SpaceX Beat Revenue by 13%. The Stock Fell Anyway. Palantir and Nvidia Held.

This article is educational only and does not recommend buying, selling, or avoiding any security.

Two trading days, one theme

Tuesday closed at a record. The S&P 500 rose 1.79% to 7,736.52, the Nasdaq Composite gained 2.59%, and the Dow closed above 54,000 for the first time. Strong earnings from Palantir and Caterpillar did some of the work. So did comments from Treasury Secretary Scott Bessent suggesting a deal that could reopen the Strait of Hormuz was within reach.

SpaceX reported its first earnings as a public company after that bell rang. Wednesday’s session showed a more selective market: the Dow added another 0.87%, but the S&P 500 finished essentially flat, the Nasdaq and Russell 2000 slipped, and market breadth was weak. Put simply, most individual stocks had a worse day than the headline numbers suggested. The equal-weight S&P 500, a version of the index that treats every company the same size instead of letting giants drive the result, fell 0.38% even as the standard index held. Investors were also waiting on the ADP jobs report and a pair of services PMI readings (surveys that ask business managers whether conditions are improving or worsening, an early read on the economy ahead of official government data), both due Wednesday.

At the center of that shift: three companies tied to the same “AI spending” theme, and three different verdicts from the market.

SpaceX’s first earnings beat came with a catch

SpaceX’s second-quarter results, its first since going public (its IPO) in June, beat expectations on paper. Revenue came in at $7.8 billion, up 92% from $4.1 billion a year earlier and roughly 13% above the $6.93 billion analysts expected. The net loss narrowed to $541 million, an improvement of $467 million from the $1.0 billion loss in the same quarter last year, with a loss of $0.09 per share against a consensus estimate of $0.26, the average of what Wall Street analysts had predicted going in. Adjusted EBITDA, a measure of how much cash the core business generates before accounting for debt costs, taxes, and non-cash items like depreciation, rose 191% to $3.5 billion.

The company’s three reporting segments told different stories. Connectivity, which includes Starlink, grew revenue 66% to $4.3 billion, with subscribers doubling year over year to 12 million. The AI segment, which includes cloud compute agreements, Grok, and X subscription revenue, grew revenue 247% to $2.6 billion. Space, the launch business, grew a more modest 29% to $962 million.

What moved the stock wasn’t the top line. It was two numbers further down the report, capital spending and share count, both of which changed faster than revenue did. More on that below. On the earnings call, Elon Musk also moved up the company’s target of $1 trillion in annual revenue to 2030, a year earlier than previously guided.

The stock fell in Tuesday’s after-hours session and kept falling all through Wednesday, closing at $108.29, down 13.6% on the day from Tuesday’s $125.33 close. That leaves it well below the $135 IPO price from June and off more than 50% from the roughly $200 all-time high it reached shortly after it began trading on June 12.

There’s also a second, unrelated pressure on the stock this week: the expiration of insider lockups on Thursday, which lets early investors and employees sell shares they’ve been restricted from selling since the IPO. More shares available for sale can weigh on a price independently of anything in the earnings report.

Why a beat can still send a stock lower

A revenue beat describes what already happened. A stock price is a bet on what happens next. When those two signals point in different directions, and one changes more than the other, the second one usually wins.

Two numbers in SpaceX’s report explain why this quarter split that way.

First, capital spending. Total capex rose from $2.8 billion a year ago to $18.4 billion this quarter, roughly 6.5 times higher. Inside that number, spending on AI infrastructure specifically went from $749 million to $15.8 billion, a jump of roughly 21 times. That’s not a rounding change to an existing budget line. It’s a new, much larger, and much less certain commitment, and it changes the question an investor is actually answering. A beat on last quarter’s revenue says the business did well last quarter. A 21-fold jump in AI capex asks whether years of that spending will convert into revenue and profit at a pace that justifies it, and one quarter isn’t enough evidence to answer that yet.

Second, share count. The number of shares used to calculate SpaceX’s loss per share roughly doubled year over year, from about 2.9 billion to about 5.9 billion, largely a function of June’s IPO and the capital it raised. Revenue grew 92% over that same period. Growth in the top line and growth in shares outstanding aren’t the same growth, and a business that keeps raising capital to fund a bigger buildout can dilute existing shareholders even while the underlying business improves.

SpaceX’s own management pushed back on the concern during the call. CFO Bret Johnsen said the company has been “efficient” with its spending, and that on the AI side it is deploying capital “in such a way that we’re getting less than a one-year payback.” That’s the company’s argument for why the capex is worth it. Whether the results bear it out is something only future quarters can answer, which is exactly the tension the market is pricing.

Neither of these facts says SpaceX made a mistake, and neither predicts what the stock does from here. What they show is that a single headline number, even a genuinely good one, is rarely the whole picture. The capex line and the share count sit a few pages deeper in the same report that produced the revenue beat, and both mattered more to Wednesday’s price action than the number that made the headline.

Same AI theme, three different verdicts

SpaceX wasn’t the only AI-adjacent name reporting this week, and it wasn’t treated the same way.

Palantir’s second-quarter report landed a day earlier and did the opposite: shares jumped 29.45% Tuesday to close at $162.66 on 93% revenue growth and a tenth consecutive quarter of raised full-year guidance. The stock opened only slightly lower Wednesday morning, slipping less than 1% in premarket trading, before giving back more ground over the full session and closing Wednesday at $158.43, down about 2.6% on the day. Even with that pullback, it held the large majority of Tuesday’s gain. Goldman Sachs raised its price target (an analyst’s estimate of where a stock should trade) to $220 from $200; the average analyst target sits near $187, about 18% above Wednesday’s close.

Nvidia moved too, closing Wednesday at $221.11, up about 4.3% from Tuesday’s close, after SpaceX named Nvidia hardware as part of its AI computing buildout. The same capex figure that dragged SpaceX’s stock down helped lift the company supplying the chips behind it.

Three companies, one AI narrative, three outcomes: rewarded for AI revenue growth (Palantir), rewarded for supplying AI infrastructure (Nvidia), and questioned for the cost of building AI infrastructure (SpaceX). “AI stock” is doing a lot of work as a label this earnings season. The businesses underneath it are not interchangeable, and neither is how the market is pricing them.

It’s also a reminder that a single company’s earnings report, however dramatic, is only one data point in a portfolio built around many. SpaceX, Palantir, and Nvidia all sit under the same “AI theme” this week, and all three moved in a different direction. A concentrated bet on any one of them would have felt very different than exposure spread across the theme as a whole.

What else is moving markets

Geopolitics is still part of this week’s story. Tuesday’s rally leaned partly on hopes for a deal that would reopen the Strait of Hormuz, a shipping route that carries a significant share of the world’s oil supply. That remains a live watch item, not a resolved one.

On the economic data side, Wednesday’s session was waiting on the ADP private payrolls report and those services PMI readings mentioned above, both scheduled for release the same day. Soft hiring signals and rising costs were cited as part of why Wednesday’s gains were narrower and more selective than Tuesday’s broad rally.

Risk and Context

SpaceX is a newly public company. Stocks with a short trading history and a small public float can carry more price volatility than more seasoned issuers, in either direction, and a double-digit move in the weeks after an IPO is not, on its own, a verdict on the underlying business. Lockup expirations, like the one this week, are a normal part of the post-IPO calendar and can add short-term selling pressure that has nothing to do with a company’s results. A single company’s earnings result, whether SpaceX’s, Palantir’s, or any other name mentioned here, does not predict how the broader market or the “AI trade” as a category will perform going forward. Geopolitical developments, including the status of the Strait of Hormuz, remain uncertain and subject to change without notice.

None of this is a signal to act on. It’s a reminder that the headline number and the full picture aren’t always the same thing. The capex line, the share count, the segment mix: that’s usually where the real story sits, a few pages past the number that makes the headline.

For the investor who reads the balance sheet, and the one who just wants to know what happened. Here at VEM, we tailor to both.