SpaceX Beat Earnings, but Its Treasury Says More Than Its P&L

SpaceX beat Wall Street's revenue and earnings expectations in its first post-IPO quarter. That is the kind of sentence that produces rallies. It should not. The first lesson in reading a corporate balance sheet is that the income statement is a story, while the cash account is a fact. SpaceX's actual facts are these: it converted billions of IPO dollars into money-market funds, left its Bitcoin position untouched, and pushed 86% of quarterly capital expenditure into AI infrastructure. The ledger remembers what the market forgets. A record IPO is not a growth trend; it is a liquidity event. On Aug. 4, SpaceX reported second-quarter revenue of $7.814 billion, up 92% year over year, against analyst expectations around $6.8 billion. Adjusted EBITDA climbed to $3.538 billion, a 191% increase, well above the $2 billion consensus. The net loss narrowed to $541 million from $1.008 billion, and the operating loss improved to $143 million from $970 million. By any operational standard, the company is executing. But the stock market does not pay a premium for the past; it prices the next set of decisions. The first decision is asset allocation. That allocation appears in the treasury schedule. SpaceX generated $85.675 billion in net proceeds from its IPO, lifting cash, equivalents, and marketable securities to nearly $100 billion. At June 30, money-market funds stood at $65.625 billion, more than triple the $21.339 billion reported at the end of 2025. Government securities classified as cash equivalents added another $4.011 billion, and marketable securities included $6.487 billion more. Together, money-market funds and government securities were worth $76.123 billion. Mapping the invisible currents of liquidity means following this allocation, not the press release. A company that chooses money-market funds over risk assets is signaling what it thinks about the next twelve months. The signal is not bullish. The Bitcoin line is even more precise. SpaceX held 18,712 BTC with a cost basis of $661 million at June 30, identical to the figures reported at Dec. 31. The filing does not disclose transactions, so offsetting buys and sells are possible. But the observable result is a treasury that moved from roughly 6.6% of its liquidity in Bitcoin at year-end to 1.1% after the IPO. The fair value of the position fell to $1.098 billion from $1.637 billion, an unrealized loss of $539 million. The position remains $437 million above cost, yet the relative allocation has collapsed. In a bull market, with a $100 billion cash pile, the decision to do nothing is a decision. Crypto investors who expected SpaceX to become a Bitcoin proxy should read that page twice. The capital that is not going into Bitcoin is going into AI. SpaceX directed $15.828 billion to AI infrastructure in Q2, more than 21 times the $749 million spent a year earlier and roughly double Q1. That was 86% of total capex of $18.369 billion and exceeded the $13.09 billion Wall Street expected for AI spending. Across the first half, AI capital expenditure reached $23.551 billion versus $3.316 billion in the prior year, nearly 83% of total investment. The build-out supports data centers, compute leasing to Google and Anthropic, and the company's own Grok and X products. AI revenue reached $2.561 billion in the quarter, so this is not purely speculative. In my experience, the CFO's payback math requires scrutiny. Bret Johnsen said contracted compute deployments are reaching payback in less than one year. That is a specific claim about specific contracts, not a segment-wide return. The AI division still produced a $1.257 billion operating loss, $1.885 billion in depreciation and amortization, and $2.178 billion in research and development spending. SpaceX has signed $14.1 billion in cloud agreements, with $6.7 billion in new contracts after the quarter. Johnsen expects capital spending to stay at similar levels through year-end. Architecture reveals the true intent: this is now an AI infrastructure company that also launches rockets. The market, however, is not only looking at the income statement. It is watching the unlock. On Aug. 6, insiders become eligible to sell about 900 million SPCX shares, worth roughly $105 billion at current prices. Tom Dunleavy, head of venture at Varys Capital, called it one of the largest lockup expirations in market history. Those shares already exist; the expiration only makes them tradable. A second block becomes eligible after Q3 earnings, and more restrictions fall on Dec. 8. By then, close to 40% of outstanding shares may be free to trade. Musk's own stake stays locked until June 2027. That is a supply schedule, and markets trade supply schedules. Short interest is already elevated. S3 Partners estimates that 95% of SPCX shares available to borrow are out on loan, and short interest is 34% of the public float. That is confirmed equity shorting. Crypto derivatives are a different signal. CoinGlass data reviewed by CryptoSlate shows SPCX futures volume reached about $6.85 billion in 24 hours, with open interest near $720 million. But open interest is two-sided: every contract has a long and a short. Rising OI does not clearly prove a directional bet. The pressure on SPCX comes from the impending supply unlock and from existing equity shorts, not from a futures book that could equally be balancing hedges. Neither pressure alone explains the selloff. The unlock sets the ceiling; AI spending sets the narrative; the conservative treasury sets the tone. Together, they create a stock caught between an ambitious capex story and a cautious cash policy. Certainty is a liability in this domain, and the market is correct to price in uncertainty. The lesson is structural. SpaceX is no longer an adolescent private company; it is a public balance sheet with an AI appetite. The first post-IPO report was designed to fund the next build-out, not to comfort shareholders. The next phase will not be decided on an earnings call. It will be decided by how many unlock holders sell into the volatility, and whether the AI contracts mature into cash flow before the short crowd rotates. Patterns repeat, but the participants change. For SPCX traders, the option-like nature of the equity short and the two-sided futures position means position sizing matters more than conviction. Survival is a function of position sizing. The consensus is often the contrarian trap. Watch the treasury, not the ticker.