Post

三份卷子已交，一份待批——加密财报季的最后一张彩票兄弟们，这周有四场开奖。 8月3日盘后 Palantir 8月4日盘后 AMD & SpaceX 8月5日盘前 Circle 加密侧的卷子已经交了三份：Coinbase、Robinhood、Tether。三份卷子，三个完全不同的剧本。有人考砸了，有人转型了，有人闷声发大财。最后一张——Circle——周四早上出分。先别急着押注。看完这三份卷子，你就知道最后一张该怎么下注了。第一份卷子：Coinbase——考砸了，但砸得“有预期” 营收12.2亿美元，同比下降18.5%，比市场预期的13亿少了8000万。每股GAAP亏损1.36美元，远高于预期的0.42美元亏损。净亏损3.59亿美元。这是连续第三个季度营收不及预期。盘后跌了约5%。但有意思的是，正式开盘后跌幅扩大到了13.9%。为什么？因为市场在等一个东西：Q3指引。 Coinbase没说Q3能反弹。没说加密交易量会回暖。没说订阅收入能稳住。当一个公司交了一份烂卷子，还不给你任何“下次会更好”的承诺——市场只能用脚投票。但Coinbase不是没亮点。现金及现金等价物86亿美元，其中支付稳定币29.6亿、货币市场基金及短期美债30.9亿。它抗得住。问题是：抗得住不等于涨得动。第二份卷子：Robinhood——转型成功，但加密凉了总营收13.1亿美元，同比增长32%，创历史纪录。净利润5.73亿美元，同比增长48%，每股收益0.62美元，远超预期的0.41美元。漂亮吗？漂亮。但拆开看：预测市场收入1.56亿美元——第一次超过了加密货币和股票交易收入。加密货币收入同比暴跌38%，只剩1亿美元。加密名义交易量400亿美元，其中Robinhood App端180亿，同比下降35%。 Robinhood用预测市场和股票交易，补上了加密的窟窿。这说明什么？说明散户还在交易，但不在加密里交易了。他们去了预测市场，去了股票期权。对Robinhood来说是好事。对加密来说，这是个危险的信号——散户的热情正在从币圈撤离。第三份卷子：Tether——熊市里唯一笑得出来的那个净营业利润15亿美元，比Q1涨了将近50%。USDT流通量1846亿美元，市场份额冲到60%以上。超额储备41.1亿美元。用户基数突破6.5亿。一个季度赚15亿美金，比多少上市公司一年赚得都多。 Tether的生意本质是什么？吃利息。美国国债收益率高，Tether拿着USDT的储备去买美债，利差就是利润。只要美联储不降息，Tether就能一直印一直赚。最讽刺的是：加密市场越熊，Tether赚得越多。因为熊市里大家更愿意持有稳定币。三份卷子看完，一个结论很清晰： Coinbase在熬，Robinhood在转，Tether在赚。最后一个——Circle——8月5日盘前交卷。市场关注什么？USDC流通量、储备收益、机构合作进展、稳定币竞争格局。最乐观情景： USDC流通量环比增长、利息收入超预期→提振DeFi流动性预期。最悲观情景： USDC继续萎缩、合规成本上升→引发市场对稳定币赛道竞争加剧的担忧。 Q1结束时USDC流通量约770亿美元，同比增长28%。Q2能不能守住这个势头？Tether已经用1846亿的USDT告诉你了——稳定币市场，赢家通吃。交易时间表（北京时间）：周一（8月3日）盘后 → Palantir（PLTR）。市场预期营收约18亿美元，同比增长80%，EPS约0.28-0.35美元。AI数据赛道，P/S ratio 61倍——高估值不允许任何失望。周二（8月4日）盘后 → AMD & SpaceX。AMD预期营收约113亿美元，同比增长约48%，EPS约1.61美元。AI芯片赛道，今年以来涨超120%。SpaceX是上市后首份财报，预期营收约69.3亿美元，EBIT亏损约15.5亿。但真正的炸弹是8月6日的9.115亿股解禁——比现有流通股还多。摩根士丹利直接说这是“最危险时刻”。周三（8月5日）盘前 → Circle（CRCL）。稳定币第一股的首份完整季度财报。操作建议（说人话）：每个财报公布前1小时，小仓位博弈可以。但切忌重仓押注单一事件。公布后，看指引，不看数字。上一轮微软、亚马逊、Meta、苹果——营收全部超预期，股价走了四个方向。差别在哪？在下一季的指引。 Coinbase跌，是因为没给指引。 Robinhood涨，是因为预测市场给了新故事。 Tether赚，是因为商业模式不需要讲故事。 Circle的财报数字不重要。重要的是：USDC的流通量是在涨还是在缩？财报季不是赌博季，是预期差收割季。你不需要猜数据，你需要猜市场对数据的反应。周四早上，Circle交卷。 $BTC $CRCL $AMD #财报观察员：本周四场开奖，Circle压轴

Tonight Palantir, tomorrow night AMD—crypto folks need to keep their eyes wide open for these two earnings reports Brothers, tonight and tomorrow night, the US stock market has two "draws." After market close on August 3, Palantir. After market close on August 4, AMD. You might think: What does this have to do with me? I don’t trade US stocks. But I want to tell you—these two earnings reports might be tied to your crypto positions on the same rope. First, Palantir. What does this company do? AI data analysis and integration platform. Simply put, it helps governments and enterprises "mine gold" from massive data using AI. Market expectations for Q2 revenue are $1.81 billion, up 80% year-over-year, EPS $0.35, up 116% year-over-year. But Wall Street is in an uproar. Cleveland Research says commercial client spending is softening, and PLTR dropped 6% in response. Baird and Oppenheimer fired back immediately, saying Palantir beats its own guidance every quarter, "the problem isn’t demand, it’s capacity constraints." How big is the disagreement? The options market prices post-earnings stock price volatility at ±9.6%. One side up 20%, the other down 20%, the casino is open. What does this have to do with crypto? If Palantir’s guidance beats expectations = accelerated enterprise AI adoption. Enterprises willing to pay more for AI software means the entire AI industry chain’s demand logic is validated. Crypto projects doing AI data analysis and smart contract auditing—their valuations will rise accordingly. If Palantir’s guidance misses = enterprise AI spending may have peaked. Then the whole AI narrative gets discounted, crypto AI tokens? They’ll be the first to lose liquidity. Now, AMD. Q1 revenue $10.25 billion, data center $5.775 billion, up 57%. Q2 guidance $11.2 billion, up 46%. Market expects EPS $1.61, up 235% year-over-year. AMD has risen 114%-130% year-to-date, crushing Nvidia’s 12%-18% gains in the same period. Analysts’ average target price is $582, with 20% upside. But the key is— the better AMD’s AI chips sell, the harder it may be for crypto miners to get GPUs. AMD is shifting focus from gaming and mining to enterprise AI and data centers. It no longer sells just single chips but packages CPU+AI accelerators+network+software into complete solutions for data centers. What does this mean? GPU capacity is shifting from "miners" to "AI giants." Bitcoin miners are already turning hardware toward AI workloads. Decentralized GPU networks Akash and Render face supply and pricing pressure from centralized AI computing power. If AMD’s earnings further validate AI chip demand, GPUs will only get tighter and more expensive. Mining rig costs rise, PoW miners’ profit margins get squeezed. But conversely—the value of computing power is also validated, strengthening the narrative for decentralized computing power rental projects. A harsh truth: Coinbase revenue dropped 18.5% year-over-year, Robinhood crypto revenue fell nearly 40%. Traditional CeFi is bleeding, but capital is seeking new outlets. AI is the only beta in current tech stocks—the whole market relies on AI. And crypto AI is the alpha within that beta. But the premise is: the traditional AI narrative must not break. Palantir and AMD’s earnings are two stress tests to check if this rope is strong. My strategic thoughts (for reference only, not investment advice): If both Palantir and AMD soar after earnings → the traditional AI narrative is reinforced, watch for crypto AI tokens (Render, Akash, io.net, etc.) to follow the rally. Coinbase just launched Render, with 24-hour trading volume hitting $340 million—liquidity is piling in this direction. If one misses expectations → the AI sector may be under pressure overall, but a big drop is actually a window to position in computing power projects. The core logic of decentralized computing power—cheap, permissionless, censorship-resistant—is most valuable when centralized computing power is expensive. If both collapse → then don’t bottom-fish yet, wait and watch the broader market. $BTC $AMD $CRCL #财报观察员：本周四场开奖，Circle压轴