Eric SJ

Attention all traders:

From now on, before making any trades, especially in the morning, I will post an intraday trend analysis for the assets we're targeting, usually BTC, ETH, and SOL.

The trading direction and analysis for the day will primarily follow this note, and the trend must remain consistent unless I explicitly indicate a direction change.

If there's a violation, I'll be sending out red envelopes for each infraction!

I discovered something quite counterintuitive

As everyone (probably) knows: Hyperliquid’s on-chain application revenue has been consistently #1, while Pump.fun has been trailing not far behind for the most part

But when I was整理 Q2 on-chain operational data, I found that:

In Pump.fun’s most recent two quarters, the protocol revenue actually surpassed Hyperliquid

This result is definitely a bit unexpected

➠ For now, I’ll set aside costs and won’t discuss them first (I’ll get to it later)

Every time I open DefiLLama, Pump.fun is always behind Hyperliquid
I didn’t expect that, when summed quarterly, #2 would end up being more aggressive than #1

_____________________________________

I compared the revenue mix further:

Pump.fun has as many as 8 revenue components, and none of them individually contributes more than 35%

But for Hyperliquid, a single transaction fee accounts for over 80%

So in terms of overall revenue scale, Pump.fun might be the 👍 (this is the most counterintuitive part of the data)
But when it comes to profit quality, Hyperliquid is the 👍 (which makes sense)

And this is just based on these two data points

📍 Quick teaser: later I’ll break down the Q2 “on-chain financial report” data

TSMC $TSMB This earnings report is strong enough

Yet the stock price isn’t celebrating—it’s under pressure instead

Master Shi, please translate it for me. What exactly counts as a “surprise”?

The issues TSMC faces now are no longer about whether “revenue can grow.” Instead, it’s about whether growth can continue to beat market expectations that are already extremely high.

What the market wants is no longer just a surprise.

It wants one surprise after another—

just like ASML, which delivered its results the day before, and this time TSMC also put out an earnings report where it’s almost impossible to find any obvious fundamental flaws:

➠ 2026 Q2 net profit of about $22.36 billion, up 23.4% quarter-over-quarter, and up about 74.3% year-over-year—clearly above market expectations
➠ Revenue of $40.2 billion, also at the upper end of the company’s prior guidance range of $39.0 billion–$40.2 billion

More importantly, as a heavy-asset manufacturing company—while building the world’s most expensive fabs, continuing to invest in the most advanced processes, and also bearing equipment depreciation, energy, labor, and overseas expansion costs—yet it still managed to deliver profit margins close to those of a software company.

That’s where TSMC is truly outrageous.

But for these AI-sector giants, the situation is actually a bit awkward:

They’ve already reached a position where everyone agrees they’re excellent.

And once “excellent” becomes a consensus, the ultimate investment returns depend less on whether it’s excellent, and more on how much it can surpass the consensus…

📍 So: when a company’s excellence has become an overwhelming consensus, is it still a good enough investment?

$BTC 63500 is support. Within the day, on the left side here, take a long position; the cost-performance ratio is quite good

For real, $BTC just got both long and short positions at the same time

Risk-reward is 2 to 1

That’s it—going to sleep. Nothing more to elaborate. Just cut losses and be done

This level at $BTC is a resistance point. I don’t think it’ll go up that smoothly.

So come on, my friends—let’s see if we can pull off both long and short for profit. In the previous post, the risk-reward ratio and the win-loss ratio were both better.

First, let’s look for a pullback.

I think we can try placing a long near the current price of $BTC

For the intraday chart, this area might be a bit on the left side, but when you look at the overall trend, the current position is exactly at the 0.382 retracement support within an uptrend

All the indicators are also relatively low, and the integer support around 64000 isn’t far

The cost-performance of going in is really quite good

(I didn’t check whether there’s any bad news out there)

ASML $ASML The earnings report sent yesterday— I think the guidance is worth reviewing first
Because in the past two days there has been a lot of controversy around AI infrastructure, but the guidance ASML provided is even more optimistic than market forecasts—including the most conservative scenario

Three key data takeaways:
1. The entire Q3 guidance range is higher than external forecasts
2. The entire full-year guidance range is higher than the market consensus
3. When revenue guidance is raised significantly, the gross margin guidance is also lifted overall

As AI-related stocks have fallen in the secondary market, skepticism about the durability of AI demand has been increasing

But now ASML not only believes market demand will remain strong—the pace of revenue recognition/realization is also accelerating

Otherwise, you wouldn’t see both the midpoint of the Q3 and full-year guidance coming in above external forecasts by double-digit percentages

More importantly, the lower bounds of both sets of guidance are also above the corresponding forecasts

And if only the Q3 guidance were unusually high, the market could still suspect the company is simply pulling forward subsequent revenue recognition

But now it’s not only that the Q3 guidance is clearly higher than forecasts—the full-year guidance is also lifted overall, and compared with the prior version (i.e., the official guidance from the previous quarter), the midpoint is directly raised by about 15.8%

📍 This suggests the change is more likely driven by a shift in full-year operating expectations rather than just a quarterly fluctuation

While the full-year revenue midpoint increases by roughly EUR 6 billion, the gross margin midpoint actually rises by 3 percentage points

At least based on the guidance results, the added revenue does not appear to significantly dilute the company’s overall profitability

➠ The specific reasons will require further breakdown of the product mix and equipment delivery-related data (I’ll try to send it later today)

But just from the guidance, ASML is not pursuing growth by “scaling up with low-profit revenue.” Instead, both revenue and profitability are being revised upward in tandem

The market is still debating whether AI demand will cool off

At ASML, they’re already discussing just how much delivery and revenue can exceed expectations...

Issuing a stablecoin is not as technically difficult as people imagine; the real difficulty is: who is willing to accept it?

The other day, I pulled some numbers—citing data from DeFi Llama—and found that there are 147 stablecoins with a market cap greater than 1M. It’s a dizzying figure, and the number is still climbing...

Comparing USDT and USDC can help us discuss this issue more effectively:

The former’s advantage over the latter is not only that its reserve size is larger. It has nearly tracked the entire historical development cycle of crypto. At this point, it has been embedded in Web3 and can be used in every part of the transaction workflow.

The essence of money is a credit credential that is mutually accepted. It solves three things: denomination, exchange, and store of value.

A supplement:

In my previous posts, I equated the model for reserve-backed stablecoins with the “issuance size × U.S. Treasury yield”

That’s wrong.

That logic may work for Tether, but it doesn’t apply when you move to the next tier

You also have to consider how much yield the issuer needs to allocate to the distribution channels in order to sustain that scale—i.e., the simplest total profit minus costs formula, especially how much remains after subtracting distribution costs

Change your thinking to this logic

For this stablecoin space, reserve yield is only gross revenue; distribution capability determines the final profit.

Circle’s on-season total revenue and reserve revenue are about $694 million, while distribution, trading, and other costs reach about $407 million

At least with Circle and $USDC , it’s not as if the money can just be issued and “lie back and earn.” Exchanges, wallets, custodians, and channel partners take a substantial portion of the economic value; revenue-sharing paid to partners and on-chain issuance trading costs are included in this figure

This matters because it shows that a stablecoin business model can’t be viewed only as “issuance size × U.S. Treasury yield.” It also depends on how much of the issuer’s revenue must be distributed to channels in order to maintain that scale

Note: Tether and USDT’s financial data have not fully disclosed cost details, but it’s widely suspected that its cost expenses are likely far lower than Circle’s

So I’d rather think of stablecoins as an internet platform business: reserve returns are just gross revenue, and distribution capability determines the final profit

Circlc only holds the right to issue USDC, but it doesn’t completely own the user entry points for USDC

The users who primarily use and interact with USDC may come from Coinbase; trading volume may come from Binance; and use cases may come from on-chain activity and a particular DeFi protocol

This is similar to traditional internet platforms. An e-commerce brand has the products, but it needs the platform, logistics, and payment networks to complete the transaction

Stablecoins work the same way: the issuer provides the “on-chain dollar” product, while exchanges, wallets, public chains, and payment platforms deliver the product to users

If an issuer lacks its own user entry points, it needs to buy distribution with revenue

(But conversely, if a stablecoin has already become the foundational liquidity for an exchange, cross-border settlement, or emerging markets, then channel partners are intrinsically dependent on it, and the issuer’s bargaining power increases significantly—this is also why I inserted a remark about Tether above)

On the $BTC nd day, within today, I will directly look at 1000 points down!

In fact, from the order book we can clearly see that during those earlier stretches of upward movement, the length of the bullish candles has been continuously diminishing.

After the initial rebound of the first segment, it entered a period of sideways consolidation and grappling, then once it broke upward again, the upward momentum was already starting to weaken.

Up until last night—does everyone remember my previous post? I said that at $ETH 1860, the shorting opportunity had a very high cost-effectiveness.

I went in—and I still lost.

Then later, I chased again at 1880. It fell to around 186 and I got taken out, but I was able to come back and break even on the profit and loss.

So this part is etched very clearly in my memory.

Now with BTC here, I found that when I woke up this morning, it pushed up to 65000.

But in this stretch, it seems to have exhausted the strength of the longs.

Short!

Using the position of Ether $ETH 1860 for a retracement has a very high cost-effectiveness!

There are currently more than a hundred stablecoins in the market, and more and more issuers keep coming in. These issuers come from a wide range of enterprise (B) and government (G) sectors.

As stablecoins further integrate into traditional finance and payments in the future, I believe there will certainly be an integration solution for stablecoins.

This solution will be able to be compatible with almost all the major stablecoins used in the market.

Otherwise, we would have to use USDT, USDC, USDS, DAI, USD1, USDe, USYC, USDG, BUIDL, PYUSD, USDYRLUSD, USDD, USDf, U, USDGO, USDTB, pUSD, GHO, USD0, YLDS, USX, EURC, TUSD, rwaUSDi, FDUSD, GUSD, apxUSD, BUSD, crvUSD, AUSD, USDai, USAT, reUSD, FRAX, EURCV, FLEXUSD, satUSD, MUST, CASH, MOVEUSD, USDF, avUSD, FRXUSD, DUSD, USDAT, DOLA, UUSD, USDA, cgUSD, LISUSD, USDon, USDSUI.......

At the same time, future issuers may also become new “entry points” for users with a certain scale, like @RobinhoodCrypto.

For example, OUSD not long ago.

Without getting into whether the authenticity of that list is reliable, just consider why it was widely discussed: because behind the entities in that list, there are almost all very large payment networks.

We can hardly imagine what it would look like once these traditional payment networks integrate stablecoins.

And this kind of “alliance” could be the early form of a [unified integration solution]

Trivia: The $USDG issued by Paxos can be used for on-chain trading (to counter memes) on the Robinhood Chain.

And thanks to the activity on the RH Chain, the trading volume of $USDG on RH Chain has surpassed its trading volume on Solana and Base in one fell swoop.

There are more than a hundred stablecoins in the market, but the distribution channels are very limited.

Traditional financial platforms are also looking for their own on-chain dollars.

In the future, stablecoin competition won’t be just single-coin competition, but:

Platform × Stablecoin × User ecosystem competition.

I think after this cycle, the stablecoin landscape will change dramatically.

I mean, get it now!

This week, if your market value isn’t 100 billion, you don’t even qualify to sit at the table and chat.

This week’s U.S. stock earnings season is officially underway.

I’ve整理ed it, and there really are quite a few individual stocks worth discussing. Broadly, they can be summarized into three main themes: AI / banks / healthcare.

1. AI infrastructure: TSMC $TSM , plus ASML and Seagate, covering three stages of the semiconductor ecosystem—semiconductor equipment, chip manufacturing, and storage.

2. Bank stocks: earnings reports are being released across the board. Representative of this is $JPM . You can use the perspective of the entire sector to look at the U.S. economy.

3. In healthcare, from Johnson & Johnson $JNJ , to UnitedHealth, to Abbott—you can trace almost the entire healthcare industry chain, from the payment side to the manufacturing side (from insurance / healthcare coverage, to pharmaceuticals, to medical devices).

So this is also a content preview. After I finish sharing my views on the earnings reports of individual stocks, I may also pull it together and review the whole picture from a broader perspective.

As for other stocks—like BlackRock and Netflix, which everyone is more familiar with—my views on those will be shared alongside as well.