Orbit: Crypto Community Feed

A few days ago, I said Elon Musk was going to sell the Tesla Shanghai factory, and the comment section said it was all debunked. That's because you don't understand Musk; his denials are even more genuine than confirmations 😂 In-depth analysis: Why Tesla Shanghai won't be sold, and why Musk can't achieve a closed loop from chips to satellites, from ground to space? On August 6, SpaceX and Tesla jointly announced the construction of a super chip factory. One of these companies has US defense contracts, and the other is a giant with 45% of global production based in Shanghai. A military aerospace company that completely excludes Chinese factors is going to deeply bind with an automaker that owns the largest single factory in China. The security reviews between the two countries can delay or kill the project. There are also three actions that are more sincere than words: First, restricting Chinese employees' access to global data. According to multiple media reports, Tesla may be separating the office systems of China and the US businesses, restricting Chinese employees from directly accessing business data and systems in other regions. Second, initiating supply chain relocation; SpaceX is completely excluding Chinese factors from its global supply chain. Third, Tesla invested $2 billion to purchase SpaceX shares. In March this year, Tesla spent $2 billion to buy SpaceX shares. Although the stake is less than 1%, the significance of this transaction lies not in the shareholding ratio but in the upgrade from business cooperation to capital connection between the two companies. So these two companies are not only deeply cooperating in business but may even merge in the future. Musk's response to the merger question was that it is not suitable to discuss on a conference call and must follow proper procedures. This is not a denial; it is not a denial. The Shanghai factory is too important, which is why it is even more dangerous. In the first half of 2026, the Shanghai factory will deliver nearly 468,000 vehicles, accounting for more than 54% of global production. Such an important asset, how could Tesla possibly give it up? But on the other hand, precisely because the Shanghai factory is so important and deeply connected to the Chinese supply chain, it becomes more sensitive in the context of Tesla and SpaceX integration. The larger the Shanghai factory, the higher the future cost of separation, and the "institutional distance" that must be maintained between it and sensitive businesses like SpaceX must also be greater. Musk's chip empire blueprint is clear and grand: use Terafab to achieve high-end chip self-supply, use SpaceX's military orders and Starlink network to build space AI infrastructure, and use Tesla's autonomous driving and robots to consume computing power. This is a closed loop from chips to satellites, from ground to space. But this closed loop has one premise: it must pass US national security review.

Recently, $BTC has been stuck at 64,000, and many people are asking where the money has gone. To put it simply: liquidity. The money hasn't disappeared; it was intercepted halfway by AI. Huang is teaming up with Blackstone and Goldman Sachs to leverage 500 billion to build AI factories. SpaceX burns 15.8 billion on AI in one quarter, Google Cloud has a backlog of 514 billion and is still calling for more — behind these massive capex figures are huge bond issuances. US Treasury yields are being pushed up, institutional funds are all buying AI bonds, so who’s left to buy BTC? This is why ETFs have had inflows for five consecutive days, BTC touched 65,000 but was slammed back down — incremental funds are being absorbed by AI, and the crypto market is all about stock competition. Looking at ETF data makes it even clearer. Last week, spot ETFs had net inflows for five consecutive days, totaling $853 million, the strongest week since April. But yesterday (August 10), there was an outflow of $144.6 million in one day, ending the five-day streak. What’s really interesting isn’t the outflow itself, but the internal differentiation of funds. Of the $853 million last week, BlackRock’s IBIT alone accounted for $693 million, 81% of the money flowing to one place. What about other ETFs? GBTC is bleeding, FBTC is bleeding, BITB is also bleeding. Yesterday it was even clearer: IBIT outflowed $53.6 million, GBTC $52 million, FBTC $40.3 million, BITB $28.4 million — but Grayscale’s Mini Trust actually had an inflow of $37.1 million. Do you see? The money isn’t leaving BTC; it’s moving from high-fee products to low-fee products. GBTC has high fees, so funds keep flowing out; Mini Trust has low fees, so funds are flowing in. This isn’t bearish on BTC; it’s institutions saving on fees. But the problem is, this internal shuffling doesn’t bring net buying, so BTC remains under pressure. $ETH is even worse than BTC, stuck around 1900, with ETH/BTC steadily declining. Last week, ETH ETFs still had net inflows of $245 million, the fifth consecutive week of inflows, but on August 10, it flipped to net outflow, breaking the momentum. Even worse is on-chain: L2 fees are drained, burning is almost meaningless, staking withdrawals can dump anytime, and selling pressure outweighs buying pressure. BTC lacks incremental funds; ETH can’t even hold the incremental funds it gets, so it often lags on the upside and leads on the downside. More trouble comes from selling pressure beyond ETFs. In Q1 this year, listed mining companies sold over 32,000 BTC, more than the total sold in all of 2025. After halving, mining rewards were cut in half; 15%-20% of hash power is operating at a loss, so miners must sell or shut down. Mining difficulty dropped 10% in June, inefficient miners are going offline, but as long as the price hovers around 64,000, marginal miners’ selling pressure won’t stop. But on the flip side, AI computing power and crypto mining are now competing for electricity and data centers. Many mining companies have shifted to AI computing power hosting. Miners sell BTC to survive while converting data centers into AI data centers to earn new revenue, which may accelerate industry consolidation — surviving miners won’t rely solely on mining income, and selling pressure may ease in the long term. Citibank is even more pessimistic, cutting its 12-month ETF net inflow forecast from 10 billion to zero — note, this isn’t an extreme assumption, but the baseline scenario. Combined with July’s ETF posting the smallest monthly inflow since launch, institutional buying is indeed slowing. Of course, it’s not all bearish. Long-term holders’ locked positions have reached 79%, a record high, meaning truly circulating supply is shrinking. Under this structure, once there’s a catalyst (like a mild CPI tomorrow night or renewed rate cut expectations), the price could rally quickly. But in the short term, BTC at 64,000 faces resistance at 65,000 and support at 62,000. ETF internal differentiation, continuous miner selling pressure, and macro liquidity drained by AI leave little ammo for a breakout. Don’t chase longs or bottom-fish blindly; wait for data to guide direction. Do you think this ETF outflow is a one-day event or the start of a trend? Let’s discuss in the comments. #现货ETF资金分化，BTC卖压仍在