The plan was to start adding short positions yesterday and complete the build within a week, so why is it dropping so fast...

Changxin placed an order at 30 yuan this morning, but since it wasn't filled, I don't plan to play anymore. However, seeing that many people on Hype are shorting or hedging, and the trading price on Hype is essentially a T+1 price discovery, I opened a position again. Currently stuck, waiting to exit before Hynix's earnings report on Wednesday.

This market situation doesn't inspire any desire to bottom-fish at all. Right now, what I want most is for BTC to spike up with a wick, then short it hard.

A brief recap of this week. This week was a bloodbath for semiconductors, and the market is starting to doubt how much longer this AI Capex cycle can last. At the same time, the momentum strategy in the US stock market is still collapsing. Simply put, the stocks that performed best and were most crowded recently are now being sold off together. I think this signal is more important than how much the index has dropped because it indicates that deleveraging is not over yet. But more than the decline, Kimi is also worth paying attention to. Because open source ≠ cheap training. We don’t know how many GPUs K3 actually used behind the scenes, nor the real training costs, so using it directly to prove the end of scaling is definitely not rigorous. But on the other hand, it’s clear: model capabilities are still improving, yet prices keep dropping. As open-source models get closer to closed-source, how long can Anthropic’s ARR growth slope be maintained? I think definitely not for long. (If it really goes public in October, that will be a judgment on AI) Looking further up, big tech Capex has also entered a very awkward phase: Growth is no longer enough; it must exceed expectations. The market has long priced in AI spending for the next few years, so if next quarter’s Capex is just in line, it might be treated as below expectations. Interestingly, this weekend, the Chinese A-share market is again releasing a flurry of positive news. Nine ChiNext companies simultaneously released half-year earnings forecasts, all positive, and basically covering the sectors hit hardest on Friday: Copper foil, chips, optical devices, optical modules, computing power services—the entire AI computing power chain is basically covered. China Reform Holdings and China Chengtong also came out to express continued increased holdings in central enterprises, tech stocks, and ETFs. In Hong Kong, the US will no longer extend the national emergency related to Hong Kong; South Korea announced a roadmap for the internationalization of the won, preparing to gradually lift restrictions on won trading and settlement. This weekend, almost all the positive news that could come out has come out. However, I remain pessimistic about next week’s performance. The biggest problem now may no longer be fundamentals, but that the rise was too fast earlier, and positions and leverage in the market are still too high. Especially with US stock momentum still collapsing, it shows that the clearing of crowded trades is far from over. Once deleveraging starts, it can’t be immediately stopped by a few positive news. Better to lie low and rest rather than mess around blindly.

The greatest truths are simple; support and resistance are truly the core of short-term trading, and stop-loss is easy to implement.

$QQQ Let's have a rebound, it's been painful

I think the opportunity for many people with Changxin's IPO might be to open positions on the Hype. IPO price is ¥8.66, corresponding to a market cap of about ¥579.2 billion. Initially, 1.471 billion shares will be circulating, about 2.2%, but the issuer can additionally allocate 1 billion shares, so less than 4% circulation, which looks very low. Think about SPCX🤡 I think the odds for over ¥3 trillion are not enough, I'll exit below 40.