07/14: The shoe drops tonight.

Today, the Shanghai Composite Index probed the bottom with a “Golden Needle” intraday dip, and after falling below the 200-day line, it completed the recovery on the very next day. The day’s positive candle body fully engulfed yesterday’s negative K-line body. Besides volume not keeping up, the overall price action has temporarily removed the technical risk of a short-term pullback. [Taoguba]
After hours, looking back, the market today was full of twists and turns. As the Korean market swung wildly up and down, it was truly nerve-wracking. The title from yesterday’s recap was “No need to panic.” After watching the opening auction this morning, I really didn’t panic either. Later, I went out to the hospital, and when I came back I saw that the account that had been in profit at the open suddenly flipped into a large loss. I’m not used to using my phone to watch the market, so I wasn’t clear about the live moves on the screen, and honestly my heart did start to race with worry.
Around midday, news that the Korean government was preparing to step in to rescue the market began to go viral, and there’s not much more to say about the afternoon trend.
Many people saw the news: yesterday, multiple top-tier quant firms came out to explain that their quant models did not include a Korea stock factor. Last night, I also saw so-called analysis articles from people in the industry claiming to prove that. So, does today’s performance count as a slap in the face? The market was already starting a repair trade from the open. Yesterday, the number of 172 limit-down stocks during the opening auction stage quickly fell to fewer than 10. Later, when the Korean stock market plunged, the A-share market followed and kept selling down into the market’s bottom zone this year. In the afternoon, the rescue-market news from Korea triggered a rebound: after the Korean market opened it stabilized, rallied, and closed with a deep “V” reversal turning red at the close. Since the Korean market opened an hour earlier than the A-shares, quants had already switched on a bulldozer-mode straight after the afternoon open, without hesitation for even a minute.
So, if you want to trade tech stocks, you really need to keep a close eye on how the Korean market moves. Every single move they make affects whether A-share quant funds are adding fuel or driving the knife.

Today’s market once again vividly demonstrated A-shares’ “anti-human” trait:
Before the open, domestic computing power received major positive news, and yet today the main force chose to smash the domestic computing chain;
Commercial space was heavily sold off yesterday, and even after-hours there were still positive messages. Many people thought today would at least see a rebound. China Satellite, which had hit the limit-up board against the trend yesterday, became the bellwether for whether the aerospace sector could repair. The result: China Satellite, with a market cap over one trillion yuan, hit the daily limit-down within one minute of the open.
When the index fell into the valley in the morning and many people felt desperate, there wasn’t even time for retail investors to think during the afternoon open—it just kicked into a rally. When many people charged in to buy the dip, tech was once again halted, and the theme switched to biopharma.
People who chased the domestic computing chain at the open today, and those who didn’t cut losses on aerospace yesterday but hoped for something—both felt awful. Those who tried to buy the dip on the tech line in the afternoon but didn’t catch the China-supply optics-module PCB—also felt awful.

From a technical perspective, today’s move is still not enough to judge a reversal of the market. The key for the next step is volume—whether the rise can be sustained. If the rebound happens on shrinking volume, the rally can only be categorized as a pullback/rebound lasting within 2 trading days. If it rises on flat volume or a modest expansion and keeps going for 3 trading days or more, then it becomes a phase rebound. Only sustained expansion in volume and big consecutive rallies for 5 trading days or more have the foundation for a reversal. Whether the market has already turned into a trend reversal is hard to judge right now. But what’s optimistic is that the three “unknowns” have landed today. One is the liquidation storm in Korean stocks; one is the performance risks at the leaders; one is the Fed’s rate-hike expectations.

A few months ago, the Korean stock market made countless A-share investors envious. Now, everything happening today should make countless people grateful they weren’t born in Korea. Some veteran A-share investors say the Korean market is essentially precisely replicating the tragedy of A-shares in 2015:疯狂杠杆 drove a crazy rally, and then management urgently stepped in to cool things off and tighten leverage. After that, countless accounts got wiped out and countless people’s assets went to zero overnight.

Yesterday, I saw professional people saying that the biggest storm in the Korean stock market would come tomorrow, and they also said this was one of the reasons why capital had fled the tech line aggressively yesterday—fearing being dragged down by Korean stocks tomorrow. Today’s deep-V rebound in Korea has caused many accounts’ margin ratios to return back above the 140% safety line. Tomorrow won’t trigger forced liquidation, which greatly eases the sell-pressure for the next day. This “unknown” can’t be said to have fully landed yet, but the expectation of the government rescue has reversed so far. It is now clear that on Thursday the Korean government will hold a special meeting.

Tonight is the final deadline for the Shenzhen Stock Exchange’s board mid-year report performance mandatory disclosure. Dongshan Precision, which has been dragging its feet for a long time, finally delivered a very impressive answer tonight, and all the talk that it would miss expectations is now gone. Multiple top brokerage firms gave a 2026 full-year attributable net profit midpoint of 3.5–3.7 billion yuan (China Merchants, CITIC, Huatai, and others). The most aggressive institutions’ highest forecast for the first-half report is also only 2.8 billion, while a steady-style forecast is 2.0 billion. The result: Dongshan did 3.0 billion. Although Yi Zhongtian is not within the scope of tonight’s mandatory disclosure, the prosperity of the industry chain has transmission effects. Dongshan Precision’s impressive “answer” is enough to clear away all doubts about top manufacturers’ performance. This “unknown” about earnings can be considered landed.

The other “unknown”—and the biggest one—is the U.S. CPI data to be released tonight. I won’t copy the specific details. It came in below market expectations, which sharply lowered the odds of rate hikes. This “unknown” is also considered halfway landed. The other half is tonight Beijing time at 22:00, when Fed Chair testifies before Congress. It’s expected there won’t be any clearly hawkish content. Iran announced tonight that it is exiting the ceasefire agreement—does that count as officially starting the fighting? But the market chose to ignore it. U.S. stock index futures were up strongly before the open, the U.S. Dollar Index fell, and gold, U.S. Treasuries, and oil all rose. There’s also an implied transmission effect here: when the dollar weakens, global liquidity tightness will ease accordingly, which is good for Asia-Pacific equities—and also good for a group of U.S. tech giants that are extremely sensitive to financing pressure.

It seems the hardest time has already passed.

Today’s moves:

In both the morning and afternoon, I only watched one thing before going out, which led to me getting hurt again today. If I had sold Huaihai Qingke in the morning, I could have made a bit of profit; when I sold in the afternoon open, I ended up losing 6%. I’m not selling because I don’t like it—I just wanted to bet on a super-node. I switched to Huafeng Technology, which had fallen even deeper. Into the close, I bought into Zhuoyi Information on the dip. This was a bet on domestic GPUs. It doesn’t do GPUs itself, but every domestic GPU depends on it for BIOS firmware. The logic is the same as Xinyuan Shares: every chip can’t do without its IP license.