The trend of $ETH is a bit awkward. Originally, because the long-short ratio is higher than BTC and SOL, if the CPI is positive, it could perform stronger than others. But it dropped last night, which probably scared off quite a few people. Because ETH's long-short ratio is 72.3%:27.7%, which is 2.61; higher than BTC's 1.6 and SOL's 2.01, the stampede last night was even more severe. If tomorrow's CPI is negative, it is expected to be another strong stampede. Hopefully, tomorrow will be positive. After all, a wave of leverage has been cleared, which is good for a rebound after the CPI.

Unexpectedly, AI has become a direct competitor to web3... AI has become so popular that even Bitcoin mining companies have stopped mining and switched to leasing to AI companies. Riot Platforms secured a $9.1 billion deal with Anthropic. After the news broke, Riot's stock surged 25% after hours. The lease is for 191 megawatts of power capacity at its Texas mining facility. If Anthropic wants to build a new data center, it would have to wait years in line for power connection. But Bitcoin mining farms already have power connected, and since mining is currently unprofitable, they are eager to find new opportunities. This is the best time to undercut them. #财报观察员：AI基建财报接力登场

I've been thinking these past two days: why did the $BTC ETF have a net inflow of $854 million last week, hitting a new high since April, yet the price didn't go up? It even turned into a net outflow yesterday. What exactly are institutions doing? Just now while driving, I suddenly thought, could it be that last week's inflow was users relocating due to the Coldcard incident? That would also explain why 80% of the funds flowed in from IBIT. If that's really the case, it means my judgment last week was wrong; maybe institutions are already taking profits now, which fits with my feeling that there's still a big downtrend ahead. The more I think about it, the more it seems to be true. #本周三CPI公布，9月加息定价会改写吗？

$OKB is really strong. Last night, $BTC dropped 2 points before the CPI data release, but OKB was basically unaffected and even rose, showing an independent trend. 1. A few days ago, after the non-farm payroll data boosted rate cut expectations, OKB surged 10%. The normal profit-taking correction from the day before yesterday has ended, and judging by the current momentum, it clearly wants to push higher again. 2. From the current perspective, as long as the price doesn't fall below 88, the overall trend remains upward. Moreover, this is an independent trend relying only on the broader market, X Layer, and the OKX ecosystem. Although it will be affected by the CPI data, it has the potential to maintain an independent trend. 3. Friends who haven't bought OKB can consider building some base positions; there will be opportunities to perform. But don't hold too heavy a position. Those who want to accumulate can choose the low points when everyone is panicking. I previously accumulated some OKB by participating in the flash earning event, so now I feel completely at ease. #财报观察员：AI基建财报接力登场

Just saw the latest Glassnode indicator, the $BTC Seller Exhaustion Index has entered the extreme exhaustion zone for the first time. This is a composite indicator measuring low volatility and high loss states. Historically, previous signals appeared at the bottom of bear markets. Although entering the extreme exhaustion zone this time doesn't guarantee it's the bear market bottom, I don't aim for the absolute lowest point when bottom-fishing, just a relative low is enough. So I just added a large position. If $BTC later remains volatile or continues to decline, but the Seller Exhaustion Index does not hit a new low, then the second time would indicate the bottom. Friends looking to bottom-fish can consider taking action now

MicroStrategy has been selling coins recently, and I see some friends are worried about MicroStrategy crashing and dragging down $BTC Actually, there’s no need to be afraid at all: 1. It’s normal for MicroStrategy to sell coins; they are an investment institution, and their profit model is buying low and selling high. The initial claim of "never selling" was just to rally more people to support the price, so don’t fully trust it. Even if they don’t sell this cycle, they will sell in the next. Moreover, they sold coins in the previous cycle a few years ago; this is not their first time selling. 2. This cycle, MicroStrategy won’t crash because they won’t be troubled by the annual 1.76 billion in interest. In fact, MicroStrategy has many options. First, issuing common stock. Their common stock issuance capacity is about 24.6 billion. Second, issuing perpetual preferred stock. Their perpetual preferred stock issuance capacity is 25 billion. Third, selling $BTC , whose total BTC value is 58.5 billion. Also, when they really stop caring about goodwill, the interest they must repay annually is only 2 trillion. Not paying other interest won’t be illegal; their contract terms are very flexible. 3. Even if MicroStrategy sells coins to repay debt, BTC’s current market cap is 1.2 trillion, with a daily spot trading volume of 30 billion and a daily derivatives trading volume of 150 billion. Such huge trading volume won’t be crashed by an annual 1.76 billion sell-off. #Strategy再卖1690枚BTC，企业财库出现分化

$BTC Is the sentiment really this bad now? Yesterday I saw BTC stagnating, and I thought maybe all the funds were waiting on the CPI. But the CPI hasn't even been released yet, and it already dropped below 64,000 last night? I guess some institutions are betting on the CPI exceeding expectations. 1. This week's market mainly depends on tomorrow's CPI data: the expectation is a year-over-year 3.42%, core 2.52%. If it's below expectations, it could mean no rate hike or even a rate cut, and BTC would go up; if it exceeds expectations, it will break below 62,000. 2. Institutions had net inflows yesterday, indicating some funds are withdrawing, probably because some institutions speculate the CPI data will exceed expectations. The current fear and greed index is 31, sentiment is positive, so retail investors probably won't run. 3. Regarding the CPI data and last week's non-farm payroll data, some friends asked if there could be falsification. It's actually quite possible, but consider this: government falsification is also to serve monetary policy. Non-farm payroll and CPI data themselves are meant to serve normal monetary functions, so even if falsified, interpreting policy from the data is still reasonable. But I don't recommend betting on the CPI in advance, because retail investors' information sources are still much worse than institutions. Don't turn investing into a game of betting on size or luck. #本周三CPI公布，9月加息定价会改写吗？

$DOS retail investors really never match up to the project team; I even suspect the project team deliberately added the wrong liquidity pool. Yesterday's $DOS airdrop had a pre-market price of 0.32u, clearly with a 4% hard sell pressure, but the project team's liquidity pool addition maneuver completely outsmarted everyone. 1. In Figure 1, this sniper bought over 20,000u at 1.4, directly cutting the sniper players; 2. Later, the order book updated, with prices fluctuating between 0.148 and 0.164. Many retail investors habitually rushed to front-run, with quite a few selling at 33u, basically absorbing 1% of the sell pressure. 3. Finally, the project team somehow managed a very low community airdrop claim rate. So when everyone was scared off by potential community sell pressure, the project directly pushed the price up to 0.4, comfortably unloading their holdings. 4. By today, the sell pressure was basically gone, and the project team started releasing positive news, coordinating a pump to continue stable selling above 0.4u. People who play altcoins are truly talented, always coming up with new strategies #Strategy再卖1690枚BTC，企业财库出现分化

The two leading mining companies together lost 850 million, both have shifted to playing with AI, hoping the $BTC bear market ends soon. #现货ETF资金回流，BTC与ETH能否接力？ In the past two days, MARA and CleanSpark have successively released their Q2 financial reports. MARA reported a net loss of 611.3 million USD, compared to a profit of 808 million USD in the same period last year; CleanSpark reported a net loss of 239.8 million USD, compared to a profit of 257 million USD last year. Revenues dropped by 27% and 30.5% respectively. The major losses came from the fair value loss of $BTC. MARA recorded 343 million USD, CleanSpark recorded 116 million USD, totaling 459 million USD. When the coin price falls, the balance sheet immediately blows up. To pay electricity bills, MARA sold 2,213 BTC in Q2, reducing inventory from the peak by 29% to 35,577 BTC. Now both companies are heavily promoting their AI infrastructure transformation. CleanSpark signed a 20-year, 6.6 billion USD lease for Sandersville, and MARA is developing 2GW power assets in Texas. If the bear market continues a bit longer, these mining companies will probably have to hand over their chips to exchange for new teams and paths. This is one of the reasons I believe the bull market still has one last dip. #比特币BIP-110分叉停滞，矿工支持不足

Brothers, I've organized the tokens to be unlocked this week for everyone, please be cautious to avoid risks: 1. The most dangerous is $YZY, the biggest bomb this week: it has a circulating market cap of 88.71 million, and this time it will unlock 35.73 million at once, accounting for 22.8% of the circulation. This is equivalent to suddenly adding 1/4 more chips to the market out of thin air. With a small market cap and such a high proportion, it will most likely crash sharply first and then decline gradually. Be careful to avoid it. 2. Next is $STRK, which will unlock 130 million tokens this time, accounting for 3.6% of circulation, amounting to 3.19 million, which can be considered a hidden bomb. Although the amount doesn't look big, the market is so bad now, and the Layer2 sector is already struggling. STRK has been declining all year without reversing the trend, so this unlock is like giving ammunition to those who want to dump. 3. The rest are relatively less risky. AVAX/APT/SEI/ARB all have relatively large market caps, with proportions below 2%. Especially $ARB, the 260,000 amount is basically negligible, and $AVAX's 15.31 million compared to a 2.79 billion market cap is like plucking a hair. Basically safe. 4. How much token unlocking affects the market actually depends on three numbers: proportion, market cap, and hype. If the proportion is greater than 5%, there is real selling pressure; if the circulating market cap is less than 1 billion, the effect is easily amplified; if the sector has no hype, it means fewer buyers to take over, so the dump will be harsher.