When will SpaceX's stock plummet? When is the best time to buy the dip?

June 12 saw what was described as the biggest IPO in history: SpaceX surged 19% on its first trading day, and its market capitalization easily flew past $2 trillion.

The FOMO (fear of missing out) feeling was immediately ignited. Two trading days later, its market cap even briefly broke through $3 trillion during the trading day.

Elon Musk’s personal wealth also reached as high as $1.45 trillion.

And then, there was no “and then.”

Over the subsequent three trading days up to today, SpaceX has dropped 23%, and it is on the verge of falling back below $2 trillion.

Why is this happening?

A company seen as a beacon of human technology and dominating headlines worldwide—how did it suddenly start to look like a pig-butchering scheme?

01

There is nothing wrong with SpaceX’s spaceflight business.

According to the S-1 prospectus, its total revenue in 2025 will be $18.7 billion, up 33% year over year; among this, the Starlink business contributes 60% of total revenue, with nearly a 50% year-over-year surge.

If that’s the case, what is the market worried about?

A comment from Prof G, a marketing professor at New York University, precisely pinpointed the issue: Starlink really is a great business, but SpaceX is asking you to ignore all the other messes.

In 2025, SpaceX’s capital expenditures were nearly $21 billion; $12.7 billion of that was used to build data centers for xAI.

The money SpaceX spends on computing power infrastructure is far more than what it spends on building rockets and satellites.

The result is that SpaceX was profitable even in 2024, with $791 million; but in 2025, despite a major growth in Starlink, it turned to a net loss of $4.94 billion.

By Q1 2026, the quarterly net loss alone has skyrocketed to $428 million.

To a certain extent, SpaceX has already become a white glove for funding Musk’s AI ambitions.

On June 16, the day the stock hit its peak, Musk suddenly announced that SpaceX would acquire the AI programming startup Cursor in an all-stock deal priced at $60 billion.

A rocket company that had just gotten a $75 billion valuation from the market—before it even had time to “warm up”—now plans to issue another $60 billion in new shares to buy an AI tool for writing code, massively diluting shareholders’ equity.

Even more suffocating: according to disclosures on June 19, the company clearly has about $100.8 billion in cash and cash equivalents on its books, yet it needs to issue debt of $20 billion to raise funds.

The purpose is: to repay the bridge loans incurred when acquiring xAI in 2022, and to supplement general corporate purposes.

This obviously abnormal and aggressive cash flow management strategy inevitably makes the market wonder: Is that $100.8 billion in cash locked up in future computing-hardware procurement, and essentially cannot be used at all?

Issuing more shares while diluting shareholders’ equity, and also issuing debt to increase financial leverage.

Such dense financing and bloodletting has created a serious crowding-out effect on marginal funds across the entire U.S. stock tech sector.

As early as June 17, trading of SpaceX’s stock options officially opened on the exchange.

In the prior few days, investors aggressively bought out-of-the-money call options. In order to remain risk-neutral, market makers were forced to buy large amounts of SpaceX’s underlying shares in the spot market.

When high-level buy demand withered away and stalled, the Delta value of the call options dropped rapidly. Market makers no longer needed to hold as many underlying shares to hedge, so they began to sell underlying shares aggressively.

At the same time, funds that felt panicked started buying put options. To hedge the risk of puts, market makers also had to short the underlying shares to balance.

This push and pull meant that liquidity in the options market not only failed to buffer the decline, but instead became a catalyst that accelerated the drop.

More importantly, to artificially push up the stock price early on, SpaceX’s publicly tradable float was deliberately compressed to be very small—creating a classic squeeze effect and leading to a forced upward surge in the days immediately after listing.

And based on the flow predictions derived from Investing.com’s fund-flow data, next month—after the release of the Q2 earnings report—up to September of this year, as many as 20% of shares held by internal employees and early investors are expected to be unlocked.

Tens of billions of dollars’ worth of shares will hit the market within two months. Who can absorb them?

At the end of the day, the stock market is like a reservoir, and the “water” (liquidity) in that pool is limited.

On June 12, the day SpaceX went public, the entire U.S. stock aerospace sector went through an extremely brutal bloodbath:

Imaging satellite company Planet Labs fell 8.8%; rocket launch company Rocket Lab dropped 10.8%; satellite communications provider EchoStar fell 11%; lunar concept stock Intuitive Machines fell 13.1%; direct-to-phone satellite company AST SpaceMobile dropped 15.5%...

To a large extent, this proves: Right now, there simply isn’t enough incremental capital in the market to support SpaceX’s enormous scale.

SpaceX’s rise is essentially being fueled by the blood of its peers.

And as more and more unlocked shares are released every month, whose blood will be sucked next?

02

So much has been said in a pessimistic tone, and SpaceX—seemingly soaring at the height of its momentum—suddenly appears to have become a target for everyone’s criticism.

At this point, many people will certainly recall that famous saying: “Be fearful when others are greedy.”

Can SpaceX reverse course?

The answer is: absolutely.

SpaceX is undoubtedly one of the greatest companies of this century. This current plunge is simply a typical “structural pain” from an IPO.

Once the bad news has run its course, everyone’s attention will automatically shift from negative headlines to its underlying fundamentals.

What does SpaceX have?

1. Starlink.

As mentioned earlier, it’s true that SpaceX has shifted from profit to loss. But it’s also true that Starlink’s earnings are currently in a breakout phase.

As of this June, Starlink’s total number of global subscription users has already surpassed 12 million.

Its user structure is shown as below.

The largest portion—the ordinary home-use segment—is mainly used in the early phase to build scale and spread out the sunk costs of rocket launches. The real source of revenue comes from aviation onboard and ocean shipping.

Starlink has, in fact, transformed into a global enterprise-level foundational network platform.

Currently, SpaceX is deploying the third-generation (Gen 3) satellites at large scale. Each satellite can provide more than 1 Tbps of downlink capacity—over 10 times that of the second generation—while uplink capacity has increased by 24 times as well.

Once Gen 3 is fully rolled out, Starlink will not only continue to capture broadband demand in more remote areas, but also be able to swallow a portion of the traditional telecom operators’ pie through direct-to-cellphone connections.

Goldman Sachs expects that by 2030, Starlink’s revenue will reach $144 billion.

2. Starship.

Take a look at the diagram: the next-generation Starship has been comprehensively reconfigured with Raptor third-generation engines. It removes complex external thermal protection ducting, and boosts booster thrust from 7,590 tons during the Flight 9 era to 8,240 tons.

Based on SpaceX’s commercial launch cost model submitted to the market, once this model reaches mass production, the per-launch cost will fall to below $20 million. The cost per kilogram to reach low Earth orbit will be as low as $10–20.

At that time, it will not only monopolize the launch market, but also cause the deployment cost of Gen 3 satellites to fall sharply.

In other words, it’s vertical integration across the industry chain—like the left hand turning the right hand.

(Next-gen Starship power system upgrade roadmap; source: Reddit)

3. Computing power leasing contracts

In May, Musk signed a computing power leasing contract with Anthropic worth $15 billion per year (totaling $45 billion). In June, he also landed a $11 billion per year deal with Google. On June 22, in the pre-market hours, he urgently announced a $6.3 billion GB300 computing power order with Reflection AI.

You could say it has been a steady stream of good news.

But Musk effectively dug his own grave. On X, he posted: “SpaceX has not promised to rent Colossus long term; if we later face a shortage of computing power, we may take it back after 180 days.”

In the contracts with Reflection AI and Anthropic, it also indeed states: “After 3 months, you may terminate with 90 days’ notice only.”

That means the company’s most important recurring revenue suddenly becomes a temporary revenue stream that could disappear at any time.

As long as SpaceX can, in its Q3 earnings report, modify the computing power leasing contracts with Anthropic, Google, and Reflection AI into “irrevocable fixed 3-year agreements,” and remove the 180-day unlock/termination threat.

Then it can prove to the market that its massive data centers really can operate as GPU-as-a-Service (GPU services provided externally, bringing stable profits).

Its valuation logic will immediately switch from a “heavy-asset manufacturing enterprise” back to a “top-tier AI cloud service provider.”

……

The market’s concerns now boil down to: SpaceX is not earning enough money to burn on AI infrastructure and rocket R&D.

But according to the above content, as well as the forward performance forecasts by Wall Street’s mainstream investment banks disclosed by Markets Insider, SpaceX’s financial performance outlook is not expected to be that bad.

So, how should you look at what comes next?

Starting in July and running through December, as a massive wave of internal shares unlocks, there will inevitably be repeated shakedowns to complete institutional share rotation.

If you see huge trading volume but the stock price does not hit fresh lows, it means long-term institutional funds are already in the market, absorbing most of the panic selling.

The two most critical events are: the full-all-employee unlock node in December and the mass production report for the Raptor 3 engines.

Once these two things are settled, any panic sentiment will dissipate.

03

Epilogue

Recently, Nobel Prize-winning economist Paul Krugman called Musk a “human Ponzi scheme.”

Indeed, at least for now, Musk seems to be treating SpaceX as his personal infinite-credit-card—his appetite is not exactly pleasant to look at.

But this does not negate the fact that Starlink makes money, Starship is iterating, and the logic of space data centers is solid both technically and commercially.

In the long run, the space economy and advanced data services remain the correct narrative for the coming decades.

As long as someone believes Starship can bring the cost of getting to space down to “cheap as cabbage,” believes AI data centers will become the world’s computing engine, and believes Starlink can create revenue in the trillions of dollars by 2030—

even when, at this time and during the upcoming unlock wave over the next few months, various media will declare at full volume that “Musk’s empire is collapsing,” there will still be large numbers of people who stand firm or wait for the right moment.

Sifting through to remove the chaff and keep the wheat is a healthy process.