SpaceX Shorts Aren't Betting Against the Business. They're Betting Against the Math.

SpaceX short sellers have accumulated roughly $15.5 billion in paper profits as SPCX falls nearly 50% from its June intraday high. Short interest reached 219.3 million shares by late July - about 34% of the tiny public float. By comparison, that dwarfs the short campaign against Tesla, the other Musk company trading publicly. The bearish position has been building nonstop since the June 12 IPO.

Crowd wisdom supposedly has a different view. Social media forums are full of "don't bet against Musk" mantras. Even Musk himself warned on X that "the survival probability of firms who maintain a significant short position in SpaceX over time is very low."

Except the shorts aren't actually betting that SpaceX goes bankrupt. They're not arguing Starlink is dying or that launch services are collapsing. The short thesis here is simpler and more mechanical than the crowd-wisdom narrative admits: an IPO structure that made only 5% of shares available for trading, a staggering unlock schedule set to flood the market starting August 6, and AI capital expenditures that look unsustainable at current revenue levels.

The real question isn't whether crowd wisdom beats the shorts. It's whether the stock's risk/reward has finally turned attractive after the selloff - or whether the bears are right to wait before anyone, on either side, gets too comfortable.

The Unlock Is the Real Story

SpaceX priced at $135 on June 11 and briefly soared to $225.64 on June 16, briefly touching a $2.66 trillion valuation. The euphoria made sense for a company that had just merged with xAI at a combined $1.25 trillion private valuation. But less than 5% of the roughly 13 billion total shares were actually available for public trading at launch.

That means the stock's price has been determined by a thin pool of roughly 640 million shares trading against retail enthusiasm, hedge fund allocations, and thematic ETF inflows. Every dollar move in that compressed float is amplified.

Starting August 6 - two days after today's earnings report - up to 911.5 million shares can become eligible for sale. That's 20% of the locked-up holdings from early institutional backers and private investors who bought in at valuations a fraction of today's price. At current levels, that pool represents roughly $116 billion in potential supply. Even if insiders sell only a fraction, the float is set to more than double its current size.

This isn't a standard 180-day lockup cliff. SpaceX's staggered unlock schedule spreads 16 different release dates across the next year. The tradable float could expand sixfold by late September as employee tranches unlock in 7% increments. By December, the full employee pool clears. Musk's own 42% stake doesn't unlock until June 2027, which is the only anchor holding back the full flood.

The structural supply overhang is why the short position keeps growing regardless of how earnings turn out. The unlock ladder exists independently of Starlink subscriber growth, launch cadence, or AI strategy. It's pure mechanics.

The Fundamentals: Starlink Carries the Thesis, AI Burns the Cash

SpaceX is no longer just a launch company with a broadband side business. After absorbing xAI earlier this year, the enterprise now operates across three segments: space, connectivity, and AI.

Starlink remains the engine that makes the entire operation financially legible. The connectivity segment generated $11.4 billion in revenue and $4.4 billion in operating income during 2025, with operating income more than doubling year over year. The business spans five distinct sub-segments - residential, business, maritime, aviation, and government - and the margin quality varies enormously across them. Aviation customers generate annual revenue approaching $300,000 per aircraft compared to roughly $960 for a residential subscriber. A shift toward higher-margin commercial and government mix is what would actually drive long-term profitability, not headline subscriber counts.

Consensus revenue for Q2 2026 sits at approximately $6.8 billion, up from $4.7 billion in Q1. That's strong growth, and the connectivity segment margin is expected around 35.9% for the quarter. The business fundamentals in the space and Starlink segments are arguably intact.

Then there's xAI. The AI unit lost $6.4 billion from operations on just $3.2 billion in revenue during 2025. In Q1 2026 alone, it posted a $2.47 billion operating loss on $818 million in revenue - representing 17% of total company revenue but burning cash at a far-disproportionate rate. AI capital expenditures reached $7.72 billion in the first quarter of 2026, compared to $12.73 billion for all of 2025. Consensus projects total company capex of roughly $48.7 billion this year, rising to approximately $118.4 billion by fiscal 2028.

The market has grown skeptical of AI spending across the board. That skepticism is warranted here because xAI's path to self-funding isn't clear yet. The company plans to scale its Grok model to "multiple trillions of parameters," which will require even more infrastructure investment. Compute contracts could exceed $25 billion annualized once fully ramped, according to recent reporting. Investors need to see when and how AI spending starts paying for itself - or whether Starlink's margins can absorb the drag indefinitely.

The Valuation Is Still Stretched

This is where the contrarian instinct runs into the numbers. The market has dropped SpaceX hard - the stock is down roughly 49% from its peak and about 15% below the $135 IPO price. That looks like a battered setup on the surface.

But a 50% decline from an IPO-driven spike doesn't make the valuation attractive. At current levels near $114, SpaceX still trades at roughly 78 times trailing sales and 36 times book value. The P/E ratio is meaningless because the company lost money on a GAAP basis. Free cash flow was negative $19.8 billion over the trailing twelve months, with operating cash flow of $7.1 billion entirely consumed by $26.9 billion in capital expenditures. Return on invested capital sits at negative 14.7%. Total debt stands at $60.5 billion against $15.9 billion in cash, giving net debt of $6.6 billion.

The IPO priced the stock as if the future was guaranteed. The selloff has partially corrected that excess, but 78 times sales for a company that hasn't yet turned a GAAP profit isn't a contrarian entry point. It's a post-euphoria reality check that hasn't finished running its course.

Price Action Shows No Exhaustion Yet

SpaceX's RSI sits at 38.3, approaching oversold territory but not there yet. The MACD line is negative at -11.8 and still trending downward. The stock has lost 28.6% over the past 20 days. There's no bear trap pattern forming, no volume drying up on the sell side, no clear sign that the bears have exhausted their conviction.

Every dollar move in the stock is worth more than $300 million to the short side. That creates squeeze risk if positive earnings surprise the market or if management provides strong AI monetization guidance. But it also means the shorts have room to add to their positions as long as the unlock overhang looms.

AInvest's aggregate signal labels SPCX a Buy, but the composite analysis score of 0.86 is lukewarm - the rating reflects the quality of the underlying Starlink and space business, not the current stock setup. The liquidity rating of 7.77 shows the shares are liquid enough to trade, but the thin float makes every move violent.

So What's the Setup?

SpaceX is reporting its first earnings today after the close. The consensus expects a loss of roughly $0.23 per share versus Q1's $1.19 loss - narrowing significantly, which would be encouraging if it holds. Revenue of $6.8 billion would beat Q1's $4.7 billion. The earnings report itself is a sideshow to the larger structural story.

The short thesis here is fundamentally sound, at least for the near term: the unlock schedule creates massive structural selling pressure, xAI's cash burn is enormous with no clear monetization timeline, and the valuation hasn't compressed enough to absorb both risks. The crowd wisdom on social media isn't the counterargument - it's noise.

The contrarian entry point for SpaceX isn't here. The stock needs to work through the unlock schedule, show AI spending discipline or a credible path to self-funding, and compress its valuation multiple further before the risk/reward tilts toward the long side. A post-unlock dip, combined with evidence that Starlink margins can cover the AI overhang and that the float has expanded to a more stable level, would be where the actual opportunity starts.

For investors already in the stock, the position carries acute two-way risk around today's earnings and the August 6 unlock. For those on the sidelines, patience is the better posture. Don't chase the bounce after earnings, and don't front-run the unlock. Wait for the structural pressure to clear, then reassess.

I'd reassess this wait posture if the unlock schedule is materially modified, if management provides concrete AI monetization timelines that change the capital expenditure curve, or if price action shows genuine selling exhaustion below the $105 area - which would be closer to where the valuation stops looking stretched and starts looking like an actual opportunity.