Post

OpenAI年化营收被爆只有约500亿美元，比市场疯传的700亿少了整整200亿。英伟达跌近3%，甲骨文跌超5%，费城半导体指数重挫3.39%。比特币从83,299美元一路砸到80,397美元，24小时内超过14万个账户被清算，6.47亿美元的多头仓位灰飞烟灭。市场在恐慌。恐慌AI泡沫要破，恐慌加密要跟着崩。但Arthur Hayes昨天发了一条推，需要你记住：“钱会印，仓会涨。” Hayes说，每一轮超级牛市启动前，市场都必须翻越一堵“担忧之墙”。 2017年，墙是区块大小之争——比特币到底能不能用？2020年，墙是新冠疫情——全球经济要完蛋。2023年，墙是FTX崩盘加央行暴力加息——加密货币都是骗子。每一次，墙后面都是印钞潮。每一次，翻过去的人赚麻了。 2026年的墙是什么？Hayes给的答案很具体：AI加密技术引发的恐慌情绪。 OpenAI向投资者披露：截至9月底，年化营收接近500亿美元。而此前市场广泛报道的数字是700亿。差了200亿美元。29%的差距。市场立刻开始算账：AI空头最担心的“循环融资”链条断了怎么办？投资者注资OpenAI，OpenAI租云服务，云厂商买芯片，芯片巨头再拿钱回投OpenAI。这条链上，OpenAI是唯一面向终端收钱的环节。它的营收不及预期，意味着终端需求覆盖不了天量基建投入。英伟达跌了，甲骨文崩了，CoreWeave狂泻8%。大多数人看到OpenAI营收不及预期，第一反应是“AI泡沫要破，赶紧跑”。 Hayes看到的是：AI收入跑得越快，泡沫破得越早；泡沫破得越早，美联储印钱的理由就越充分。这不是鸡汤。这是对美联储和财政部结构性行为的判断。美国10年期国债收益率还在5.25%附近，30年期在5.69%，数十年高位。财政部要还的债越来越多，利率越高，压力越大。中期选举只剩几周，汽油价格因为霍尔木兹海峡的持续封锁居高不下——布伦特原油还在101美元以上。战争在烧钱，AI泡沫在需要兜底，选举在逼着政客做点什么。三股力量，指向同一个出口：流动性。中金最新的研报说得很直白：美联储最终降息幅度或超出市场预期，美元宽松交易可能在短期回归。 CME FedWatch显示，10月加息概率已经从一周前的70%暴跌到18%。市场还在纠结“加不加息”，但真正的老OG已经在赌“什么时候开始印”。比特币现在卡在82,000美元。下方支撑80,500-81,500，上方阻力86,700。短期确实颠簸。但Hayes的框架告诉你：每一次宏观利空释放，都是印钞预期的燃料。 AI的“鬼故事”吓崩了科技股，但吓出来的恐慌，最终会变成美联储宽松的借口。 $BTC $ETH $NVDA

BTC dropped from 82,000 to 80,000, did your position move? Or do you simply have no "movement" rules? If you are holding positions based on feeling, then today's article might be more valuable than your take-profit or stop-loss orders. On October 9th, whale jasonleo posted a message late at night. He didn't call for a long or short. He did something more ruthless: he publicly shared his "escape route." First, his core judgment: BTC has currently retraced about 5% from its high. Sounds small, but what's the background? The 30-year US Treasury yield is approaching 5.7%, the 10-year around 5.3%, rate hike expectations are heating up again, oil prices and inflation pressures are rising together. Multiple bearish factors stacked, yet BTC only fell 5%. jasonleo said a very key sentence: the price's reaction to news itself is information. In the past, any random macro bearish news could make BTC drop over 10%, but this time, with multiple bearish factors hitting, the price didn't crash, indicating the trend has resilience. So his conclusion is: he doesn't think BTC will directly fall to 78,000 or even 74,000, and continues to hold long positions. But what’s really worth copying from him is not his conclusion, but his rules. He broke down his position management into three clear tiers: 🟢 BTC stays above 79,000 → Do nothing. Keep holding longs, the trend is intact. 🟡 BTC falls below 79,000 but does not close the daily candle below 78,000 → Start reducing positions. Not fully exiting, just lowering risk first. 🔴 Daily close below 78,000 → Exit all remaining longs, stay flat and wait for the next entry. Note two details. First, he uses the daily closing price, not intraday spikes. Intraday dips below don’t count; only the close matters. Second, his average long entry price is near 78,000, which is his cost line and also his bull-bear dividing line. This is not prediction, this is preset. He wrote down in advance "what to do under what conditions." He’s not panicking at 79,000 because the rules are already set. He won’t hesitate below 78,000 because the rules are already set. And most people? They panic when it falls, get greedy when it rises, and end up making decisions based on emotions. The difference between whales and retail traders has never been who predicts better, but who has rules. You don’t need to copy his position size. His capital, cost, and leverage are completely different from yours. But you can learn his underlying logic: condition-triggered decision-making. Set three numbers in advance: observation level, reduction level, exit level. Execute when reached, do nothing if not. $BTC $CL $BZ #9月FOMC纪要公布，多数官员倾向再加息