The 100 Billion Unlock: Why SpaceX's Post-IPO Balance Sheet Is the Real Earnings Report

On August 4, SpaceX published its first quarterly earnings report as a public company. Wall Street saw revenue growth of 92 percent and an EBITDA beat that surprised nearly everyone. The company then watched its stock face the twin pressures of accelerated AI spending and a 100 billion dollar unlock. The crowd sees a moon; I see a model. The model begins not with revenue but with liquidity. Math does not care about your conviction. The report contains a balance sheet, not just an income statement. At a time when many crypto treasuries are signaling conviction by adding digital assets, SpaceX did the opposite. It took the largest IPO in market history, converted almost all of it into money-market funds and government securities, and left its 18,712 Bitcoin untouched. That should be the headline for crypto investors, not the earnings beat. The absence of movement is itself a signal. In the chaos, look for the invariant. The invariant here is not SpaceX's love of Bitcoin. It is SpaceX's need for near-zero-volatility cash to build AI infrastructure. This is a story about how post-IPO corporate balance sheets have been rewritten by the AI compute cycle. It is also a story about how the crypto market has begun to price a SpaceX equity narrative through SPCX futures while the company itself behaves less like a Bitcoin proxy and more like an investment-grade data center operator with a small digital-asset footnote. Context: From Rebellion to Compliance To understand why this quarter matters, you have to remember what SpaceX represented to crypto investors before the IPO. For years, Elon Musk's companies were treated as unofficial Bitcoin proxies. Tesla bought Bitcoin, SpaceX reportedly held Bitcoin, and the narrative was less about corporate treasury management and more about a philosophical bet on a borderless monetary asset. The May IPO filing rewrote that narrative. It confirmed a Bitcoin position, but it also introduced a new pricing framework: one where SpaceX's value would be driven by X payments, AI compute, and satellite infrastructure rather than by a digital asset pile. Now the first quarterly report after the IPO has made that framework concrete. SpaceX reported second quarter revenue of 7.814 billion dollars, a 92 percent jump year over year and well above the roughly 6.8 billion Wall Street expected. Adjusted EBITDA climbed 191 percent to 3.538 billion dollars, far above the roughly two billion dollar consensus. The net loss narrowed to 541 million dollars from 1.008 billion a year earlier. Operating loss improved to 143 million from 970 million. On the surface, these are the numbers a newly public company wants to show its institutional holders. But the real signal is further down in the filing. Core: The Balance Sheet Is the Model Revenue growth is an outcome. Balance sheet allocation is a choice. And SpaceX used its first quarter as a public company to make a very deliberate choice about capital. The company generated 85.675 billion dollars in net IPO proceeds. That cash helped push total cash, cash equivalents, and marketable securities to nearly 100 billion dollars. At June 30, SpaceX held 65.625 billion dollars in money-market funds, more than triple the 21.339 billion reported at the end of 2025. It also held 4.011 billion in government securities classified as cash equivalents and another 6.487 billion in marketable securities. Together, the money-market funds and government securities were worth 76.123 billion dollars. Let me pause here. I have spent the last decade watching crypto treasuries. Most of them talk about conviction. They publish supposed proof of reserves and announce their average cost basis. When Bitcoin falls, they say the strategy does not change. SpaceX's answer is different. It does not need to repeat mantras because its behavior is the message. The company took a historic IPO windfall and put nearly all of it into instruments that could be liquidated the same day. That is not a lack of conviction in Bitcoin. That is corporate finance in an interest rate environment where yield-bearing cash is a strategic asset. Bitcoin as a Zero-Cost Option SpaceX's reported Bitcoin position did not grow alongside its liquidity. The company held 18,712 BTC with a cost basis of 661 million dollars at June 30. That matches the amount and cost basis reported at the end of 2025. The filing does not disclose transaction-level activity, so purchases and sales could have occurred during the period and offset one another before quarter end. But the balance-sheet snapshot tells us what matters: SpaceX had every incentive to add Bitcoin if it believed in a Bitcoin treasury strategy, and it chose not to. The fair value of that Bitcoin position fell to 1.098 billion dollars from 1.637 billion during the first half, producing a 539 million dollar unrealized loss. Despite the decline, the position remained 437 million dollars above its reported cost. It represented about 1.1 percent of SpaceX's cash, cash equivalents, and marketable securities after the IPO, down from approximately 6.6 percent at year end. In my own risk models, I treat Bitcoin on a corporate balance sheet as a zero-cost call option. The cost basis is already marked. The downside is visible. The upside is uncertain but asymmetric. SpaceX could have increased that option premium when the stock was falling. It did not. Instead, it added tens of billions to money funds. That tells me the company views Bitcoin as a residual asset, not as a strategic reserve. The narrative of SpaceX as a public Bitcoin proxy should finally be buried. What remains is a minor treasury footnote. The AI Compute Machine The real capital allocation story is artificial intelligence infrastructure. SpaceX directed 15.828 billion dollars into AI infrastructure during the second quarter, more than 21 times the 749 million spent a year earlier and about double the amount deployed in the first quarter. The AI outlay accounted for 86 percent of total quarterly capital expenditure of 18.369 billion dollars. It exceeded Wall Street's 13.09 billion dollar estimate for AI spending, even as total capex came in slightly below the 18.58 billion expected. Across the first half, AI capital expenditure climbed to 23.551 billion dollars from 3.316 billion a year earlier. That represented nearly 83 percent of total investment of 28.476 billion during the period. These numbers are not a rhetorical commitment. SpaceX is building large-scale data centers and leasing high-demand computing capacity to major technology companies. It is also developing its own AI models and coding products. The investment has already begun to produce commercial returns. AI revenue reached 2.561 billion dollars during the quarter, driven by computing agreements with customers including Google and Anthropic, alongside revenue from Grok and X subscriptions. I have been building capital-flow models since the 2017 ICO cycle. In those days, people audited whitepapers by counting tokens and debating emission curves. Now the metric is contracted compute. SpaceX's chief financial officer, Bret Johnsen, said the contracted compute deployments have payback periods of less than one year. That means the company can recover equipment costs faster than it recovers costs on launch sites and satellite infrastructure. That is a powerful claim. But it comes with an important caveat: the estimate applies to specific compute contracts rather than the AI division as a whole. The segment still recorded a 1.257 billion dollar operating loss, alongside 1.885 billion dollars in depreciation and amortization and 2.178 billion dollars in research and development spending. I have audited enough infrastructure projects to know that payback period escalation is the most dangerous phrase in capital markets. A one-year payback on a specific contract is not the same as a one-year payback on a data center. The reason SpaceX is willing to spend so much is not because the economics are proven. It is because the demand for AI compute is so urgent that customers are signing contracts before the power is even connected. SpaceX has signed 14.1 billion dollars in cloud-services agreements and added another 6.7 billion in contracts after the quarter ended. That is a forward revenue backlog of 20.8 billion. But it is also a commitment to deliver compute, which means the capital expenditure is not optional. The company said capital spending will remain at a similar level during the final two quarters of the year. AI infrastructure will continue to absorb most of the investment budget. This is where the market gets nervous. The earnings beat was strong, but the forward statement is an invitation to model an escalating burn. AI capex is not like launch capex. It is more like a utility build-out. The revenue follows only after the excess heat has been converted into signed contracts. If demand falters, the depreciation alone will create a drag. That is the structural concern hidden inside the excellent quarter. Why SpaceX Is Hoarding Money Funds Instead of Bitcoin Let me be direct. The decision to hold 65.625 billion dollars in money-market funds is not a statement about Bitcoin. It is a statement about financial structure. After an IPO of this size, a company has obligations to underwriters, shareholders, and potential acquirers. It does not have the luxury of waiting for Bitcoin's next halving cycle. It has a balance sheet that must behave like a bank. Money-market funds yield something close to a risk-free rate. Government securities are the only asset that can be pledged as collateral in almost any finance transaction. Bitcoin cannot do that in the institutional world. Not yet. Narratives are liquid; truth is solid. The truth of the SpaceX balance sheet is that capital is a tool, not a belief. If you want to know what SpaceX thinks about Bitcoin, do not listen to Elon Musk's social media. Look at the percentage of total assets allocated to Bitcoin. It went from roughly 6.6 percent to 1.1 percent in six months. That is not a treasury team waiting for the right moment. That is a treasury team using the IPO to solve the covariance problem. The company no longer wants the value of its stock to move with Bitcoin because the AI business needs stable financing to buy GPUs and power contracts. The Unlock Geometry and the Short Crowd The earnings beat failed to ease concerns about two separate but connected pressures. The first is the scale of future AI spending. The second is the imminent increase in tradable SPCX shares. On August 6, insiders become eligible to sell approximately 900 million shares worth about 105 billion dollars at current prices. Tom Dunleavy of Varys Capital said the release would rank among the largest lockup expirations in market history. The shares already exist. But the expiration can sharply expand supply if employees and early investors choose to sell. The August 6 expiration is only the first stage. Another block becomes eligible after SpaceX reports third-quarter earnings, and additional restrictions expire on December 8. By then, about 40 percent of outstanding shares could be freely tradable. Elon Musk's stake remains locked until June 2027. That creates a strange dynamic: the stock can absorb increased supply, but the founder's shares still hang over the market as a longer-dated event. Short sellers had already positioned for weakness. S3 Partners estimated that 95 percent of SPCX shares available to borrow were out on loan, while short interest reached 34 percent of the public float. That is a crowded trade. A crowded short trade is not the same as a bearish verdict. It is a mechanical condition. If supply arrives and shares begin to fall, the shorts feel validated. But if insiders decide not to sell, part of that crowded short must cover. The lockup expiry is not a simple destruction event. It is an optionality event. The Crypto Derivatives View Expectations of heightened volatility were also visible in crypto derivatives markets. CoinGlass data reviewed by CryptoSlate showed SPCX futures volume and open interest reaching their highest levels since the contracts launched. In the last 24 hours, trading volume climbed to about 6.85 billion dollars, while open interest approached 720 million dollars. The first instinct is to call this a bearish signal. That is wrong. Rising open interest does not show whether traders are predominantly bullish or bearish because each futures contract includes both a long and a short position. This is where I need to be precise. The equity short data from S3 is a confirmed directional bet. Borrowed shares have to be returned, and the short seller profits only if the price falls. The crypto derivatives data is different. It is a two-sided expression of expected volatility. A trader could be buying SPCX futures because they believe the lockup will trigger a short squeeze. Another trader could be selling SPCX futures because they believe the unlock will flood the market. The open interest is simply a record of disagreement. In the chaos, look for the invariant. The invariant in this market is not direction. It is the structure of forced events. The selloff in SPCX should be connected to both pressures without claiming that either one alone caused it. The AI capex story creates a natural discount in the equity market because the investment phase is longer than the revenue curve. The unlock creates a supply question. Together, they form a confidence gap. The market does not need to choose between a bearish AI case and a bearish unlock case. Both are expressions of the same underlying fear: that SpaceX is entering a period where its balance sheet will be exposed to execution risk and secondary supply. But the crypto derivatives market shows that this uncertainty is also a trading opportunity. A futures contract on a newly public company with crowded shorts and an imminent lockup is one of the purest volatility expression vehicles available. It is not a vote on Musk. It is a vote on the calendar. Contrarian: The Lockup Is Not the Crash the Crowd Thinks It Is The consensus bear case is simple. 900 million shares become eligible. Short interest is already high. The stock is going to fall. That narrative is liquid. The truth is more solid. Lockup expirations do not force anyone to sell. They create permission to sell. Psychologically, that changes the calculus for insiders. But financially, the shares have been part of the float calculation in dark pools and private secondary markets long before the IPO. The question is not whether stock enters the market. It is whether the new buyers have already been waiting. I also think the shorts are underestimating the possibility of a negative feedback loop. With 95 percent of available shares out on loan, the borrow is nearly exhausted. If the stock begins to rally after the unlock because insiders hold and institutional buyers step in, short sellers will have to buy back at an illiquid moment. The lockup would become a catalyst for a squeeze, not a crash. The crowd sees a moon; I see a model. The model says that when supply is known and positioning is crowded, the trade is not obvious. The trade is fragile. The deeper contrarian angle is about what SpaceX is becoming. The market is treating SpaceX like a high-growth technology company with a bloated cost base. But the balance sheet says something else. The AI infrastructure division is being built with committed contracts. The cloud-services backlog covers a meaningful portion of future revenue. This is not a company spending blindly. It is a company converting its post-IPO liquidity into contracted compute, and the payback period on specific contracts is less than one year. That is not speculation. That is industrial finance. In my most recent research on the convergence of AI and crypto, I keep returning to a simple insight: AI agents will need autonomous financial systems. They will need to pay for compute, energy, and data without human approvals. Space is becoming the first company that will sit at that intersection through its own infrastructure. It is not building a blockchain. It is building the compute layer that will eventually host on-chain AI services. Crypto investors who fixate on the Bitcoin line are missing the larger transition. The future ledger will be powered by companies that own physical data centers, not by memecoins. What the Market Is Not Pricing The market is not pricing the long-term energy question. SpaceX is buying power, building data centers, and signing contracts with Google and Anthropic. That is the same thesis that Bitcoin miners have been using for years. But SpaceX has a more powerful tool: 100 billion dollars of cash and a brand that can negotiate with utilities and governments. When SpaceX enters a region looking for 500 megawatts of power, it does not have the same financing constraints as a small Bitcoin miner. It can pay for transmission upgrades. It can build its own substations. It can wait through the permitting process because the balance sheet can absorb delay. This is where the crypto market should be paying attention. The AI compute narrative is not a separate story from Bitcoin. It is a competition for the same underlying resource: electricity scarcity. Bitcoin miners have transformed themselves into AI infrastructure providers as a bear market escape plan. That strategy just received a new rival in SpaceX. The company's AI capex is larger than most mining companies' entire market capitalizations. The cost of entry is no longer just a mining rig. It is a hyperscale data center with a government-contract mindset. The implications for Bitcoin are not necessarily bearish. The fact that SpaceX is willing to spend billions on AI compute instead of Bitcoin means it believes the marginal dollar is better spent on compute than on coins. That is a fundamental signal about where the economic value of energy is being directed. But it also means Bitcoin miners who are moving into AI will face stiffer competition. They will have to justify their costs not against other miners but against a company with a nearly 100 billion dollar treasury. That is a very different game. The Philosophy of the Boring Balance Sheet Solitude is the price of clear vision. In a market that screams for narrative, the most uncomfortable truth is often the quietest one. SpaceX has published a balance sheet that is almost boring. Money-market funds, government securities, a Bitcoin position that did not change, and a massive bet on AI infrastructure. It could have bought Bitcoin with the IPO proceeds. It could have avoided the volatility of money funds. It did not. The boring balance sheet is a sign that SpaceX has moved from the rebellion stage of crypto to the compliance stage. I have written before about how the 2024 ETF approval changed the narrative from rebellion to compliance. This quarter is the first time a single company has expressed that transition in its financial statements. The IPO gave SpaceX access to institutional capital. That capital demands predictability. Bitcoin, by nature, is unpredictable. So SpaceX kept the Bitcoin it already had and placed the new capital into instruments that the bond market will accept. This is not a rejection of the Bitcoin dream. It is an acknowledgment that the corporate form has no place for dreams. Math does not care about your conviction. The Takeaway The next quarter will not be decided by the earnings beat or the lockup expiration. It will be decided by the next balance sheet snapshot. If SpaceX's cash grows, if the AI backlog continues to expand, and if the Bitcoin position remains unchanged at 18,712 BTC, then the narrative will become fully institutionalized. The company will be a utility for AI compute with a small crypto asset on its books. For crypto investors, the question is not whether SpaceX is bullish for Bitcoin. The question is whether the market can learn to price a balance sheet that no longer needs a Bitcoin narrative. SPCX futures already show us the beginning of a new pricing mechanism. The equity market is still trying to model AI capex. The crypto derivatives market is trying to model volatility. Both are looking at the same company and seeing different stories. That divergence will not last. Once the lockup passes and the AI division's cost structure becomes clearer, the market will converge on a common valuation framework. When that happens, the Bitcoin line will likely be dismissed as a rounding error. What will remain is a new category: a publicly traded infrastructure company that sits between crypto, AI, and space. I am watching the cash line, not the hashtags. The story of SpaceX for the next decade is not about the moon. It is about the machine under the balance sheet. The moon is a destination. The balance sheet is the launch vehicle. Narratives are liquid; truth is solid. The truth of this quarter is that SpaceX took a record IPO and turned it into the most disciplined AI infrastructure bet in modern financial history. The rest is noise. One final thought for the next inevitable price spike: do not confuse a futures rally with a fundamental change. The crowd may see a moon. The model sees a company converting liquidity into compute. In the chaos, look for the invariant. The invariant is 18,712 Bitcoin. It did not move. That is the quietest and most important signal in this report.