The old year’s snake hides its sharpness, the new year’s horse leaps to reveal true strength

The last segment of the Year of the Snake’s candlestick pattern hides its sharp edge, and the first batch of chips in the Year of the Horse reveals its true form. The Year of the Snake hides its edge in subtlety, while the Year of the Horse gallops freely in the vast sky. [Taogu Ba]

The final K-line of the Snake Year and the first batch of chips in the Horse Year have already quietly emerged. Who is the “Thousand-Mile Horse,” and who is the “Dark Horse”? In the Snake Year, one must hide the edge and lie in wait; in the Horse Year, it’s about “recognizing the horse, choosing the horse, riding the horse.”

The Thousand-Mile Horse looks at logic; the Dark Horse looks at sentiment; and the wild horse? It’s about courage.

One: Sentiment is a thermometer

On Monday, the index opened high and moved higher, pushing 4,300 red plates. Don’t worry about how individual stocks performed, but the face was in place. The strongest was SD2’s large one-word order. Starting Tuesday, divergence appeared, but the main force still played its part. Those that needed to be sealed off were sealed, and last night, with seedream5.0 fermenting, sentiment reached consensus. Today, new and old stocks are both winning—who runs faster?

On February 11 close, the space continued to return to a height of 4. The last high was 8 on January 29, in Baiyin Nonferrous. The next day, Hunan Gold’s attempt failed, dropping the height to 5. Later, Hengdian Film & TV, Hangdian Shares, Minbao Optoelectronics, Hanjian Heshan all hovered around 4. On Monday’s close, they hit 5, then Baituan started a second wave.

Two: Ladder thinking

Hey, Baituan, Baituan, this chart looks familiar, but what role does it play? Here’s a thought—just my random analysis~

First, what kind of stock is it? Crossing? Leading dragon? Or contrarian group? Or is it crossing and grouping?

From a thematic perspective, it belongs to the broad price increase category, with a bit of thematic grouping. Its birth was triggered by the ice point. The first wave was at the end of January—1.28, 1.29, 1.30. After the index plunged on the 2nd, it stabilized. Baituan moved sideways, and after the index weakened again, it rose again. Now, the market’s profit effect is still along the broad price increase line, and it has that upward potential, so it’s a combination of grouping + theme + ice point.

What kind of stock is this? Looking further ahead, I think it’s a connector or a stepping stone. If we consider the post-Lunar New Year profit wave, we shouldn’t expect it to go too high; it just needs to serve as a benchmark or a stepping stone—call it what you will, cannon fodder.

It’s like the chart I drew below—hope you can understand? Or you can call it a main force testing the waters.

Three: Thematic focus—emphasizing the huge computing power demand for tokens (as yesterday’s institutions mentioned, not my words)

The logic from institutions is this: the trend of token consumption and large model charging is inevitable. Therefore, token consumption will naturally explode, possibly exceeding early-year expectations. Take ByteDance as an example: the planned daily token consumption was to increase from 40 trillion to 400 trillion (a tenfold increase), but with the popularization of high-frequency tools and multimodal applications, the target might be too low.

First, driving factors: after the proliferation of Clawd Bot-like tools, users need to pay for API calls to large models (no free API), and multimodal models (like Seedance 2.0) further amplify consumption.

Second, charging logic: large model providers, under pressure from computing costs (such as ByteDance’s Seedance 2.0), have broken the “free tradition.” In the future, API charging will become mainstream.

Corresponding targets include: Wangsu Science & Technology (accelerated services), Runze + Dawei + Dongyangguang (ByteDance chain), Century Internet + Data Port (Alibaba chain), Kehua Data (Tencent).

A quick note on Wangsu: after reducing holdings last night, it started to be shared across platforms at 9:15 this morning with an update on CDN cloud service price hikes. Some key points:

CEO Matthew Prince said that the shift to AI and agents is a “fundamental reshaping of the internet,” and Cloudflare is becoming the core platform and network for AI agents to run and flow through. This positions the company from a traditional CDN and security provider to an “edge operating system” for the AI era. Cloudflare’s excellent performance provides a benchmark for market re-evaluation of the edge computing track. As a domestic leader in edge computing, Wangsu has over 2,800 edge nodes worldwide and is continuously upgrading its new-generation edge AI platform, building a full-chain capability from model access, inference optimization, to scene implementation. This is highly similar to Cloudflare’s strategy of supporting AI inference through global edge networks.

So, savor it.

What I want to convey is not just the rise and fall, but independent thinking: on February 4, Dawei Technology hit the limit down and now has three consecutive boards; Litong Electronics hit the limit down but has risen nearly 20 yuan; Wangsu Technology was engulfed in a shadow candle pattern, then hit new highs.

Four: Thematic focus—what I say, not what institutions say.

This week’s hot topic is about the dream. Since yesterday afternoon, stocks in the back row started to plunge. If Monday’s perspective A: only the first board that fermented was 360, then yesterday afternoon, it was only 0, not 36. Perspective B: Monday’s clear AI application repair, but the new favorites emerging at low levels didn’t push high again, like Blue, Zhejiang Wen, Tiandi, Lio, even if funds used Tiandi to pry the floor, it didn’t boost sentiment. It’s basically “new stocks hot, old stocks cold,” but the root cause is insufficient volume to support sector effects.

Today’s expectation is naturally that divergence will intensify, as seen in the early trading.

Specifically, the number of stocks opening with a one-word pattern has decreased—only four. After opening, Hengdian Film & TV, which showed abnormal movement yesterday, moved to a correction, indicating sentiment can’t support a main rise.

In my model, this is a tide that needs to ebb and exhaust itself. The entry point might be tomorrow afternoon or the day after morning.

I’ve heard this everywhere: the content I outlined last night is fermenting during the Spring Festival.

And just now, I saw a screenshot from Kazi Ge, which relates to the following:

DeepSeek’s new model is now in grayscale testing, bringing multiple core upgrades: greatly expanded context capacity—token length from 128K to 1 million, nearly tenfold increase, easily handling long texts like the Three-Body Trilogy.

Knowledge base fully updated: data cutoff extended to May 2025 (original was July 2024), indicating the model is likely a completely new base version (rumored to be DeepSeek-V4, subject to official announcement).

Five: Market challenges—quantitative model changes

At close, the Shanghai Composite remained red, but the ChiNext was already battered. The main area of damage was tech stocks, like Zhongji and XinYi, while resource stocks rose.

The contradiction now is that even if you see the big picture and the sectors correctly, you probably won’t make money. For example, today’s strong chemical sector, but what about the so-called red gem Polari?

In the future, I can only switch to a new mode: using emotional nodes to identify stocks near the floor in sectors (bad but not sealed), which is very counterintuitive. But what if 2026 really turns out to be this kind of gameplay for a year?

I plan to follow quant strategies, and I’m considering a candidate—Dongfang Guoxin. If it crashes tomorrow, I’ll consider entering.

Six: Your concern—rocket launch

I can reassure you that today’s launch was only a splash test, very successful. The next step is the network recovery~

But I have a question: the official launch today, why didn’t they push JuLi to the limit yesterday?

Seven: Thought question—consider two stocks from JuLi Sui’s perspective.

A: What’s the expectation for Changfei Optical Fiber tomorrow? B: What’s the expectation for Litong Electronics the day after tomorrow?

Look at me! It’s been a while since I made you think. For those who haven’t gone home for the New Year, speak up. As usual, whoever mentions me gets 10,000 points.

Thanks to the eight retail investors who boosted my last article, keeping it top quality—I love you all. @WenWeiFuBiYiLongFeng@ChaoShortLieYaoZhe@WuSu330@TouYunMuYuan007@YouGeGuaizai@DK7686@ZhuanYiDaBo@WoHuiFaGuangYeHuiFei

The last article only had five of us tipping, okay, I remember you five. I won’t let you down! @YouZiXiBaby@JinRiHeKaFeiMa@Lclx@ChenYouDi@WenDingZengYiYouRiZiYou

That’s it. Thought question—don’t forget, if you want to learn, think about it. If not, go to sleep~

Like it or not, I’ll shoot you if you don’t.