Bitcoin fell from 108350 to 91500 after December 18, with a drop of nearly 17000 points. The lowest point of the pullback last week was 91500, rebounding to 98950 early this morning, with an increase of nearly 7500 points, a rebound of 44%. It still needs to rebound around 17.8% to reach a strong resistance level of 102000. This is a crucial point for attacking 110000 recently, and if it breaks through and stabilizes, a huge pump will be opened above.

Ethereum fell to a low of 3100 after December 18, with a drop of 1000 points. It rebounded to 3642, an increase of nearly 550 points, a rebound of 55%. There is still a 6% increase to reach a strong resistance level near 3720, which constrains the breakthrough of the 4000 mark. 3660-3720 is the fluctuation zone for attacking 4000, which needs some grinding. The long position around 3300 can at least take 75% profit here.

SOL fell to a low of around 174 after December 18, with a drop of 90 points. It rebounded from 174 to 219.5 early this morning, with an increase of nearly 46 points, a rebound of 51%. So, according to the golden ratio, around 219 is an important point for short-term long position profit-taking. There is still a 10% increase to reach a critical resistance level near 229. If it breaks through and stabilizes here, the door to 257 will be wide open. Why 257 instead of 250? 257 is the watershed for attacking 300; if it stabilizes at 257, 300 will be in sight.

The profit effect of altcoins has been good these days. Today, I mainly talk about how I have been playing on-chain and in the secondary market with an 80%+ win rate recently.

First, talk about the secondary market: Recently, when playing coins in CEX exchanges, check the increases from high to low, look at each coin, and observe the strength of different coins on the 4-hour to 1-hour charts.

(1) Find the few with the strongest momentum among those with the highest increases.

(2) Compare one by one.

If the highest increase today is 15%+, look at the coins with increases of 6%-8% below, see which are new coins, have low market value, and show some upward space on the 4-hour chart (or have turned up, with previous K-line showing signs of reduced volume and not yet reaching resistance).

Pick two or three of these two types of coins, enter more, take profit at the resistance level for half, and then continue to observe the entire gain list. Now the secondary market is playing small-level reshuffling, which is quite comfortable.

If most coins reach the resistance level, you can continue to take profit on part of them. At this point, the weak coins observed earlier can be shorted for hedging. This way, even if the market drops, your overall account balance may still rise. (Strong coins drop less and rebound fiercely, while weak coins drop more and rebound less.)

Let's talk about on-chain again.

(1) Find those that have stabilized after the first wave of sell-off. Do not compete in PVP with scraping knives and stabbing swords.

Projects with low market value that have just started investing costs to operate (preferably with rumors of some well-known endorsement, etc.) examples: Aimonica, ONC, MetaV, TAOCAT

(2) Hot leading projects with strong community consensus.

No matter how high the market value of these projects is, they are still more cost-effective and practical than VC coins in CEX! Examples: KOMA, VIRTUAL, Ai16z

(3) Anti-dip coins.

When the leading coins fluctuate, observe which coins in the same track are more resistant to dips or recover quickly when large holders or a massive amount of chips are sold. Examples: Aicell (recovers quickly), Fai (anti-dip).

In the early days of AI agents, I still believe they were all manipulated by humans, but then I thought, the greatest significance of AI agents is, how do you protect your rights with 'AI'? The reason the inscriptions were popular at that time is that they can be said to be 'what everyone is fighting for together.' And the vague AI that does not point to anyone has, to some extent, increased the imagination space for valuation, even though many AI projects use technologies or products that are garbage.