摸鱼局长

Super Micro Computer suddenly released a quarterly earnings preview after the market closed last night. Revenue came in more than 20% above its previous guidance, and the stock jumped around 15% in after-hours trading.
This company assembles AI servers using Nvidia chipsets. Its shares had fallen nearly 30% over the past two weeks, and yesterday’s close landed right on the 1-year moving average.

The company’s explanation was straightforward: orders for AI platforms had increased, and gross margin was also notably higher than its previous guidance.
That said, gross margin is still only a little over 10%. The old problem of selling more but making slim profits hasn’t changed.
These are unaudited preliminary figures; the full earnings report is due on November 3.

Binance’s SMCIUSDT contract has already made its move, and is now around 38.
When the U.S. market opens tonight, first watch whether 33 holds—that was yesterday’s intraday high. If the gap gets filled, it would suggest the market isn’t buying the news.
Above that, 39 to 40 is where the stock fell from in mid-October, and the 50-day moving average is also overhead. It would need to get back above that level to recover the losses from the past two weeks.

What I care more about is whether this preview can lift the entire AI server chain. Tonight, we can also keep an eye on whether $NVDA follows along.

$SMCI
#美股 #AI servers

When news of a U.S.-Iran peace plan broke last night, Bitcoin jumped more than $1,000 in an hour, reaching a high of 83,450.
It gave back more than half the gains not long after, and was hovering around 82,800 in the early Asian session.

The 80,700 level I wrote about yesterday hasn’t been tested again over the past two days. Bitcoin also only briefly moved above the 20-day moving average before falling back.
The daily close was just below the moving average, so the breakout doesn’t count yet.

For the next 1–3 days, keep an eye on two levels.
First, the upside: last night’s high of 83,450, with the 20-day moving average just below it. A close above that level would confirm the rebound is holding.
On the downside, watch around 82,200. The pullback last night and two dips early this morning both found support there.

In the medium term, the focus is still on this range.
The top is 84,300, a level Bitcoin failed to break several times from Monday to Wednesday last week. A move above it could open the door to a retest of the late-September highs.
The bottom is 80,700, which held on three dips around last weekend. If it gives way, Bitcoin may have to retest the 80,000 mark.

A single news headline was enough to push prices higher, showing that market participants are willing to move up, but buyers chasing the rally didn’t follow through.
I’m leaning toward a strong, range-bound market. Today, the first thing to watch is whether 82,200 holds. If it does, a retest of 83,450 becomes more likely.

$BTC
#比特币 #Market Analysis

General Motors reported its third-quarter results today. They came in better than expected, yet the stock plunged at the U.S. market open.
Revenue and profit both beat Wall Street expectations. The company also announced a new $20 billion share buyback, roughly 30% of its current market cap.

The problem was its full-year guidance.
The profit range narrowed, but the midpoint wasn’t raised—and came in slightly below analysts’ expectations.
The company said memory chip and logistics costs have already climbed to the upper end of its previous estimates. The AI-driven scramble for memory is starting to show up in automakers’ income statements.

The buyback provides a floor, but there’s no timetable. Short-term investors are more focused on whether profits can keep growing next year.
The selloff pushed the stock below its 200-day moving average. Its intraday low landed right around $76.50, near the low from late September.
If it holds here, expectations around the buyback could gradually lend support. A recovery above $80.70—the 200-day moving average—would signal that the stock is repairing. If it breaks below, the next support is the trading range from July.

Tesla reports after the market closes Wednesday, offering a point of comparison: are cost pressures specific to GM, or affecting the entire auto industry?

$GM
#美股 #General Motors

Ethereum has spent the week grinding around the 2,400 level. It can’t break higher, but it hasn’t made a new low either.
It’s even weaker than BTC: the ETH/BTC exchange rate has been sliding all month, as investors prefer to stay in Bitcoin.

For the short term, look at the 4-hour chart.
Resistance is near 2,445. It was rejected there twice, on Saturday and yesterday.
First support is at 2,370. Today, two retests held above this level.

For the medium term, look at the daily chart.
Resistance is at 2,550, where both the 20-day and 50-day moving averages are weighing on price.
Support is at 2,320. Price dipped to this level twice—last Thursday and early Monday—and was bought back up both times, forming a small double bottom.

Overall, the outlook is weak and range-bound.
A move back above 2,445 would open the door to a short-term test of the moving-average band at 2,550. The pullback from the September high won’t be over until the daily chart can hold above 2,550.
If 2,320 gives way, the next meaningful support may not come until near the yearly moving average.

ETH leveraged long positions on Bitfinex remain at their highest level since 2022, with no reduction in recent days.
As long as support holds, that provides confidence. But if 2,320 breaks, those positions could fuel a cascade of stop-losses.

$ETH
#以太坊 #MarketAnalysis

BNB’s 37th quarterly burn was completed on Monday evening, removing about 1.64 million tokens worth nearly $1.2 billion at the time.
The amount was roughly the same as last quarter, and the price barely moved in the hours around the announcement.

That was to be expected.
The automatic burn is calculated using a formula based on the token price and number of blocks produced. The market already has a good idea of when burns will happen and roughly how much will be burned.
Its effect on price is more like a gradual tightening of supply, so it’s unlikely to trigger a big rally on the day.

What stands out more is its relative resilience.
Since its early October peak, BNB has fallen by roughly half as much as SOL and XRP.

Key support lies around 700–706. BNB found a floor there during two pullbacks in September and again early Monday. If it breaks below that range, the next level to watch is around 680, near the end of August.
On the upside, 750 is the first level to watch, with both the 20-day and 50-day moving averages weighing on the price there. A move back above that level would suggest the current pullback is over. Above that, the next target is 810, a level BNB failed to break through twice in late September and early October.

For now, funds are flowing into BTC while altcoins are broadly weak. BNB will most likely continue to track the broader market, trading in a range between 700 and 750.
Over the next few days, the key question is whether 700 holds.

$BNB
#BNB销毁 #CryptoMarket

Netflix reports Q3 earnings at 4 a.m. tomorrow. Its stock is already down about a quarter this year.
The company stopped reporting quarterly subscriber numbers a while ago, so this time the market is focused on advertising and guidance for next quarter.

Netflix laid off about 1,000 employees last week, and its top content executive and film chief have also left in recent weeks.
Co-CEO Sarandos admitted last month that growth wasn’t as fast as he’d hoped. Viewing hours rose just 2% in the first half of the year.
In September, several major banks downgraded the stock, all citing concerns that users’ time is shifting to YouTube.

Analysts’ expectations for Q3 revenue and profit are basically in line with the company’s guidance, so there’s little suspense there.
What will really move the stock is whether ad revenue is on track to double for the full year, and whether Q4 guidance is revised downward.
In July, guidance came in below expectations, and the stock fell to around $65 the next day.

From what I can see, the current price already reflects a lot of pessimism. As long as guidance holds steady, there’s plenty of room for a rebound.
If guidance is cut again, the July low is the key level to watch.

$NFLX
#美股 #NetflixEarnings

The “Black Monday” everyone worried about on Monday never came; the Nasdaq actually closed up a little over 1%.
Bitcoin made a brief push higher yesterday during the day, but turned lower after US markets opened. It returned to around 81,400 in early Asian trading.

For the past two weeks, BTC has mostly churned between 80,000 and 84,000.
The daily price is below the 20-day moving average and sitting right on the 50-day moving average. It hasn’t picked a direction yet.

For the next 1–3 days, keep an eye on two levels.
Above, there’s the 83,150 area, where yesterday’s high and the 20-day moving average both sit. It needs to break above that level for the rebound to gain traction.
Below, there’s around 80,700. From late Friday night through yesterday morning, the price fell to this level three times and was bought up each time.

For the medium term, look at the wider range.
Above, there’s the area around 87,300, which price failed to clear twice in late September and early October. Without a breakout, the broader trend remains range-bound.
Below, there’s last Thursday’s low of 79,800—the 80,000 level. If it breaks decisively, the next substantial support is back in the mid-September range.

I’m leaning toward a sideways market with a slight downside bias. BTC didn’t keep up with the US stock rebound, and buying pressure still isn’t strong enough.
Over the next couple of days, watch whether US stocks can keep rising tonight and whether 80,700 holds.

$BTC
#比特币 #Market Analysis

Retail traders using wallet apps to trade perpetuals on Hyperliquid have collectively lost around $280 million.
That’s the figure calculated by Hans, co-founder of Hyperdash, after tracking six wallet apps, including Phantom and MetaMask.

More than half of the money lost was actually spent on fees.
These apps add their own builder fee on top of the platform’s fees. Each trade may not seem like much.
But a typical account only deposits $30 to $50, and users keep opening and closing positions. Users of one app traded volumes worth hundreds of times their initial capital, and their money was gradually eaten away.

Three months later, only one in seven new users was still trading.

For $HYPE , there are two sides to this.
Wallet apps have indeed brought in trading volume and fees, and that revenue is directly tied to HYPE buybacks.
But if users don’t stick around, you have to keep spending money to acquire new ones. That kind of growth isn’t very sustainable.

HYPE has been weakening all month. Going forward, I’ll be paying closer attention to whether the platform can sustain its monthly fee revenue.
When trading perpetuals with a small amount of capital, fees and how often you open positions often have a bigger impact on your results than getting the direction right.

$HYPE
#Hyperliquid #Perpetual Trading

Rumors that Tesla and SpaceX might merge have been gaining traction on X over the past couple of days.
It started with a report last Wednesday by Barclays analyst Dan Levy, who said an announcement could come as early as before Tesla’s shareholder meeting on December 6. The most likely scenario, he said, would be SpaceX acquiring Tesla in a stock-for-stock deal.
Based on market caps of about $2.1 trillion and $1.5 trillion, respectively, each Tesla share would be worth roughly 0.7 SpaceX shares, before any premium.

But the stock has continued to weaken.
TSLA rose about 4.7% on October 2, the day delivery figures beat expectations, but closed at $354.72 on the 16th, giving back nearly all of those gains. It’s down about 21% year to date.
The CNN Fear & Greed Index fell to 18 last Friday, putting the broader U.S. stock market in extreme fear territory.

For now, the merger is only an analyst’s speculation, and neither company has made an official announcement. In this climate, rumors alone are unlikely to prop up the stock.
The real test is this week’s two earnings calls. SpaceX goes first at 4:30 a.m. Beijing time on Wednesday, followed by Tesla at 5:30 a.m. Thursday.
The market expects Tesla to report third-quarter revenue of around $28 billion and adjusted earnings per share of $0.44, down more than 10% year over year.
I’ll be watching closely to see how Musk responds to the merger rumors and whether automotive gross margins continue to decline.

$TSLA
#特斯拉 #USStocks

Solana’s biggest-ever consensus upgrade, Alpenglow, is scheduled to go live on mainnet at around 10:50 p.m. Beijing time on October 21.
It will replace TowerBFT, which Solana has used since genesis. According to official figures, finality time will drop from several seconds to around 150 milliseconds.
The testnet and devnet completed their migrations at the end of September, and Anza posted last weekend that everything is ready for mainnet.

The token’s price has barely reacted.
SOL closed near $120 on October 6, then spent most of last week moving between $100 and $109, hitting a low of $100.43 on the 18th.
BTC’s share of the total market cap is now above 60%, with capital flowing into Bitcoin. It’s hard for a single positive catalyst to lift altcoin prices in this environment.

I see this upgrade more as a test.
A smooth transition would simply preserve the status quo, while any hiccup could turn into a negative catalyst—especially given Solana’s history of outages.
The next things to watch are whether the migration is completed in one go on the evening of the 21st, and whether finality times remain near the official target in the days that follow.

$SOL
#Solana #CryptoMarket

U.S. bank stocks have fallen for six straight weeks. The last time the KBW Bank Index posted a weekly losing streak this long was in 2008.
From September 4 to October 16, the bank index fell 10.8%, while the S&P 500 fell just 0.8% over the same period.

Bank of America is down about 21% from its August closing high. Even relatively resilient JPMorgan has pulled back around 7%.
Earnings reports from major banks came out one after another last week, with mixed results. Fixed-income trading weighed on performance, and share prices failed to find a floor.

In theory, rate hikes are good for banks’ net interest margins. But after the Fed raised rates on September 16, the bank index actually fell faster.
Barclays’ data shows that the bank index underperformed the S&P on 18 of the 21 trading days after the hike—a first since the index was launched in 1992.
The market is worried that rates will stay higher for longer, slowing lending, increasing bad loans, and raising deposit costs. Those pressures are outweighing the extra income banks earn from wider margins.

Banks lend to the real economy, so when they diverge this much from the broader market, I’d rather see it as an early warning signal of economic stress.
As long as rate-hike expectations remain firm, a steep decline alone is unlikely to bring investors back.
The next key date is the Fed’s October 28 policy meeting. The focus will be on whether its language remains hawkish.

$BAC $JPM
#美股 #BankStocks

Predict hasn’t even launched its token yet, but it’s already starting to take market share from Polymarket!

Recently, @predictdotfun has been putting in a performance in the prediction market sector that’s definitely worth watching.

In the niche market for predicting whether BTC will rise or fall over a five-minute period, Predict’s market share has already surpassed Polymarket’s.

It’s only a niche, but taking market share from Polymarket already shows that Predict is a competitive player.

What’s more, Predict is backed by the Binance ecosystem, integrated with the Binance App and Binance Wallet, and has been mentioned publicly by CZ several times.

Right now, what everyone is probably looking forward to most is Predict’s TGE.

Aster’s performance after launch left a strong impression on many people, so naturally the market has high expectations for Predict too.

But I think Predict’s real potential lies in the prediction market sector itself.

From short-term BTC predictions to esports, sports, and U.S. stocks, if it can continue attracting users and trading volume, its future growth potential is definitely worth looking forward to.

As for how far the token can go after launch, that will depend on its valuation and market demand.

With the year winding down, Predict’s TGE is worth keeping a close eye on.

#Predict #BNBChain #Polymarket

ETH leveraged long positions on Bitfinex have climbed to nearly four times their level at the start of the month.

According to Bitfinex’s public data, ETH long positions there stood at around 94,000 on October 1. They’re now at about 368,000—the highest level since August 2022.
At a price of around $2,420, that’s close to $900 million in total.

Over the same period, ETH fell from around $2,700 to just above $2,400, a drop of more than 10%.
On October 7, ETH fell 4.6% in a single day, and long positions there increased by about 150,000 that day.

Many traders have long treated Bitfinex’s leveraged funds as a barometer of what big players are doing. This time, they’ve clearly been adding to their positions as prices fall.

My take is that this buying pressure is bullish, but it’s leveraged.
If prices stabilize, it could provide support.
But if ETH breaks below its previous low, these long positions could also become fuel for a cascade of stop-losses.

Next, watch how risk sentiment develops after the U.S. stock market opens on Monday, and whether ETH can hold above its October 15 low of $2,318.

$ETH

#以太坊 #CryptoMarket

TSMC just delivered a nearly flawless third-quarter report, but its stock hasn’t exactly rewarded it.

Q3 revenue came in at $46.72 billion, up 41.2% year over year and above the high end of its own guidance.
Fourth-quarter guidance is $52.3 billion to $54.3 billion, and its full-year growth forecast has been raised to around 44%.

Meanwhile, the Philadelphia Semiconductor Index fell for an eighth straight day on Friday, its longest losing streak since 2008.
TSMC’s U.S.-listed ADR is down about 15% from its October 5 close.

The market isn’t worried about current orders. It’s worried about how long AI data centers can keep expanding on borrowed money, especially with inflation still running hot and interest rates unlikely to fall.

One comment on the earnings call that I think is easy to overlook:
Management said consumer end-market demand is under pressure and that it will take a more cautious view of 2027.

I’m inclined to see this downturn as a repricing of valuations, rather than a sign that AI demand has peaked.
When the strongest industry leader falls along with the rest of the sector, it’s worth taking a closer look—but don’t rush to assume that cheap means the bottom is in.

Big Tech earnings are coming thick and fast. Whether they signal any easing in AI capital spending will be key to what happens next.

$TSM $NVDA

#美股 #AIChips

Charles Schwab manages more than $13 trillion in client assets, and now all eligible clients can buy and sell Bitcoin and Ethereum directly.

This isn’t a pilot—it’s a full launch.

CEO Rick Wurster made it clear at the October 15 fall business update: “All clients can now access Schwab Crypto.”

Clients can currently trade $BTC and $ETH , with a transaction fee of about 0.75%. The service is available on the website, app, and thinkorswim.
Transfers of existing holdings are still rolling out in phases, and SOL, AVAX, and LINK are also planned for listing before the end of the year.
It’s not yet available in some states, including New York and Louisiana.

I think this is more worth watching than the day-to-day price swings.
Ordinary investors can now buy crypto alongside their familiar brokerage accounts, without having to open a separate crypto account.

With easier access, there’ll be less friction for long-term capital to enter the market.
Prices may not respond immediately in the short term, but at the infrastructure level, crypto has taken another step toward the heart of traditional finance.

Would you rather buy crypto through a brokerage or keep using an exchange?

$BTC $ETH

#比特币 #CryptoMarket

The latest holdings of U.S. Vice President Vance—actually quite suitable for ordinary people to study

At first glance, it looks like there are plenty of assets, but once you break it down, the logic is very simple:
QQQ 24%
SPY 24%
DIA 24%
These three broad-market index ETFs together make up 72%
The remaining allocation is split among long-term U.S. Treasuries, gold, Bitcoin, Rumble, crude oil, and other assets

I think this portfolio has 3 noteworthy points:
① The core allocation is highly concentrated
QQQ handles tech growth, SPY covers the S&P 500, and DIA leans more toward traditional blue chips. These three ETFs basically form the “base” of the entire portfolio

② Small allocations serve different functions
Gold and U.S. Treasuries are more defensive, while Bitcoin and Rumble provide higher upside/elasticity; crude oil adds exposure to another asset class

③ What truly affects portfolio performance is still the 72% core allocation
Many people, when they see the holdings, will focus on the more attention-grabbing parts like Bitcoin and individual stocks first. But based on position sizing, the index ETFs are what ultimately determine overall returns

For ordinary investors, you may not need to copy his exact proportions, but this way of thinking is worth considering:
Put most of your allocation into core assets you believe in for the long term, and then use smaller allocations to participate in other opportunities
In the end, what investment success often comes down to is the position structure, discipline, and holding period

What do you think of this setup of “72% broad-market ETFs + 28% other assets”?

If you only know QQQ, you may not realize there are so many different ways to play the Nasdaq 100

I’ve compiled 9 mainstream Nasdaq-100-related ETFs (data standardized to 2026/7/31) and compared them across size, fees, 1-year performance, and strategy characteristics

1. The most classic
QQQ
Assets under management $452.8B, expense ratio 0.18%, 1-year return +34.03%
— The core tracking tool for the Nasdaq 100

2. Low-fee choice
QQQM
Assets under management $97.1B, expense ratio 0.15%, 1-year return +34.20%
— For long-term allocation, it’s worth paying special attention

3. Leveraged tools
TQQQ (3x bull)
1-year return +47.34%
QLD (2x bull)
1-year return +31.50%
Leverage amplifies both gains and volatility. It’s more suitable for short-term trading or investors with clear risk tolerance

4. Cash-flow strategies
QQQI
1-year return +16.01%
QYLD
1-year return +19.45%
They enhance returns via options strategies, but in a bull market they may give up some upside

5. Hedging tools
SQQQ
Nasdaq-100 -3x inverse ETF, 1-year return -46.68%
Primarily used for short-term hedging; not suitable as a long-term investment

6. Diversified weighting
QQEW / QQQE
By using equal-weighting to reduce concentration in large tech stocks

Even though they’re all “Nasdaq 100 ETFs,” the underlying investment logic is completely different

Understanding the product matters more than chasing price appreciation

The above is for information sharing only and does not constitute investment advice

Want to generate cash flow long-term from U.S. stocks, without having to pick a bunch of individual stocks yourself?
Here are 5 dividend ETFs to get to know first.

1. VIG | Vanguard Dividend Appreciation ETF
Dividend yield is about 1.54%, expense ratio is 0.04%
If you like holding high-quality companies that consistently increase their dividends over the long run, you can take a look at this one.

2. SCHD | Schwab U.S. Dividend Equity ETF
Dividend yield is about 3.13%, expense ratio is 0.06%
Among the five, it has the highest dividend yield. It balances dividends, company quality, and fundamentals—many income-focused investors pay attention to it.

3. VYM | Vanguard High Dividend Yield ETF
Dividend yield is about 2.29%, expense ratio is 0.04%
Holdings are very diversified, mainly covering large, high-dividend U.S. companies. It’s suitable for people who want a simple, low-cost way to invest for income.

4. DGRO | iShares Core Dividend Growth ETF
Dividend yield is about 1.89%, expense ratio is 0.08%
It focuses on companies that can sustainably increase dividends, making it a good fit for long-term holding and reinvesting dividends.

5. CGDV | Capital Group Dividend Value ETF
Dividend yield is about 1.18%, expense ratio is 0.33%
This is an actively managed ETF. The fund manager will actively select companies with dividend potential and value characteristics.

So which one is a better fit?
- Want higher cash flow right now: SCHD, VYM
- Want long-term dividend growth: VIG, DGRO
- Like active management: CGDV

The biggest appeal of dividend ETFs for everyday investors is simple:
By buying one fund, you gain access to a basket of companies that will keep sending you money.

Reinvest the dividends you receive—if you have enough time, you’ll gradually feel the cash flow and compounding power getting stronger.

The above is for information sharing only and does not constitute investment advice.

Latest 13F Filings from 8 Top Institutions: Where Is the Money Moving?

I compiled a chart summarizing these institutions’ Q2 reshuffling of their U.S. stock holdings. A few moves are definitely worth watching:

Berkshire: Alphabet’s position at the end of the quarter is about $37.8 billion. It continues to add to Google, while also increasing stakes in Delta Air Lines and D.R. Horton.

Tiger Global: Increased positions in AMD and SpaceX, while trimming Google, Nvidia, and Meta.

Appaloosa: Added to Amazon, Broadcom, Uber, and CoreWeave, continuing to bet on AI computing power and cloud computing.

Bridgewater: Clearly increased allocations to energy and utilities.

Pine Capital: Concentrated buys in ASML, Applied Materials, and Seagate, continuing to move closer to the semiconductor equipment and storage supply chain.

When you look at these institutions together, one money trail is becoming increasingly clear:

AI chips → semiconductor equipment → storage → data centers → power

Their AI strategy has already begun to spread outward across the industry chain, both upstream and downstream.

Google, AMD, TSMC, ASML, Applied Materials, storage, power…

Going forward, I personally plan to pay more attention to these areas: semiconductor equipment, data centers, and power.

The reason is simple: the more money gets poured into AI, the more it ultimately has to flow into chips, servers, data halls, and electricity.

You can already see early signs in this round of 13F filings—so I’ll keep watching to see how this line of investment develops next.