Crypto KOL Community & Market Trends

Shocking! Japanese art master Yayoi Kusama dies at 97
Japanese avant-garde artist and Order of Culture recipient Yayoi Kusama, known worldwide for her iconic pumpkin🎃 works
died of multiple organ failure at a hospital in Tokyo at the age of 97
The news was just confirmed today, sending shockwaves through the international art world
The master who defined a unique artistic style with polka dots, infinity mirror rooms, and pumpkin works leaves behind many unforgettable classics
🕯️

Nvidia earnings night: It initially fell 3% after hours, then rose 5%.
What a thrilling night. Let's go point by point—
One: Q2 revenue was $96.2 billion, up 106% YoY, with data center revenue at $89 billion, up 117% YoY.
It beat expectations—really beat them. The market expected $92.38 billion; it delivered $96.2 billion.
Then it fell 3% after hours.
“Beat expectations” is no longer enough for this generation of investors. You either need to “massively beat expectations” or “surpass the imagination”—there is no middle ground. This market has a bigger appetite than AI.
Two: What really triggered the 5% after-hours surge was the CFO’s 70% growth guidance for fiscal 2028.
What did analysts expect? 44%.
For the first time in its history, Nvidia gave guidance a year ahead, directly lifting expectations from 44% to 70%.
And Jensen Huang’s exact words were: “We have never before given a full-year financial performance outlook.”
The first exception was an aggressive one.
Three: Jensen Huang’s exact words: “If there were no supply constraints, that number would be much higher.”
70% is not the demand ceiling; it is the capacity ceiling.
What is real demand? 100%+.
A Bernstein analyst said this 70% implies “an increase of $200 billion compared with the previous outlook.”
In plain English: It’s not that AI is faltering; TSMC simply can’t keep up.
Four: On the same day, “Big Short” investor Burry did something inexplicable.
He bought NVDA call options as a hedge while increasing his NVDA short position.
The short position now exceeds 21% of the portfolio.
He said: “I’m not trying to make money here.” —This is risk management, not a reversal in his view.
Afraid it will rise while also afraid it will fall. This move is even more conflicted than Nvidia’s stock price.
Five: Nvidia’s narrative is changing.
From “Can it still beat expectations?” to “When will the capacity bottleneck be resolved?”
The former is a short-term trade; the latter is a long-term sector.
Vera Rubin is now fully in production, with each gigawatt of deployed computing capacity representing a $40 billion revenue opportunity. The top five customers’ capex is expected to reach $1.3 trillion next year.
The main AI infrastructure trend is far from over. #英伟达财报周 #老用户1BTC回归礼 #事件合约赛享1%交易量奖励 $NVDA $BTC $ETH

$BB ‌ has flipped bullish on the 1H 👀
Price broke above 0.0100 and is holding above the 0.00975 Supertrend. Still, the chart shows a risk warning, so I’d keep the setup conservative.
Entry: 0.0100–0.0102
SL: 0.00970
TP1: 0.01055
TP2: 0.0110
TP3: 0.0115
Key support: 0.00975–0.0100
Resistance: 0.01055
Bias: Bullish above 0.00975
A clean break above 0.01055 could open the next leg. If 0.00975 fails, I’d step back.
Nice structure, but with the warning shown on the chart, risk management matters even more here. 📈
Not financial advice. Manage risk.
#BTCPullbackto79000 #CandyDrop1BTCForOldUsers

It has to come back again when it comes to NFTs.

Although altcoins are highly volatile, Bitcoin and Ethereum are truly strong! The sideways-to-down movement has almost fully worked off the 4H indicators! If there isn’t any particularly bearish news, most likely it will keep moving sideways and then print another bullish candle that takes out everyone who opened shorts here; those who got out here or are waiting for a pullback will just have to keep waiting. All I can say is, this is damn impressive—I finally get it, and everything is crystal clear!

One dollar, multiple uses. After all the calls, it’s finally here. DeFi has it, and tokenized U.S. stocks do too!
In DeFi, a single ETH can work in three places at once: staking, LP, and lending. Stocks in a brokerage account can only do one thing: sit there.
That’s how traditional brokerages work. You buy a stock, and it just sits there. You can’t use it for LP, collateralize it, or lend it out. The only thing you can do is open the app and check whether it’s up or down today.

AI agents need fuel. $TIREA is the economic energy layer powering inference, tools, memory, data and machine-to-machine coordination.

After looking at quite a few projects combining AI with Web3, there’s one issue I’ve always felt is very practical. Many projects talk about giving AI access to real-world data. But in practice, where the data comes from and how to verify whether it’s authentic mostly remain at the conceptual stage.

TRUMP at $2.1—would you dare buy the dip?
Look at the surface first: violent pumps and dumps, with retail traders repeatedly being fleeced.
It started from $1.4 last week, surged to $3.68 in three days, then fell back to $2.1 in consecutive bearish candles. It is still up 50% on the weekly chart, but the daily chart already shows “lower highs + lower lows.” The candlesticks tell you: huge volume with long upper wicks + consecutive bearish candles—the pump is over, so don’t catch a falling knife.
First: rumor-driven pumping, followed by a slap in the face from the son—a textbook case of “buy the rumor, sell the news.”
On August 19, Trump met crypto industry representatives at the White House and said he would advance digital asset legislation. The market interpreted this as a “policy shift,” and TRUMP rose 160% in one week.
Then what happened? Someone spread the rumor that “Trump is going to launch a new coin,” and FOMO drove the price directly from $2.4 to $3.68. Eric Trump then publicly denied it: “This is a scam.” The price crashed directly from $3.68 back to $2.1.
Second: 990k wallets are down $3.8 billion, while the team is selling.
Nansen data shows that approximately 990k wallets have suffered cumulative losses of $3.8 billion on TRUMP. Meanwhile, Trump-related entities such as CIC Digital disclosed extremely high revenue—the president is making money while retail traders foot the bill.
Total supply: 1 billion; circulating supply: only 251 million (25%)
Insiders control approximately 80% of the supply, with about 900k tokens unlocking daily (at $2.1 = $1.9 million/day)
The next major unlock is on September 18: approximately 28.7 million tokens
Third: the technicals tell you that $2.1 is not the bottom—it is “mid-slope.”
From $3.68 downward, the 4H chart clearly shows “lower highs + lower lows.” Declining volume indicates that buyers chasing the rally are fading, which is not a healthy “pullback accumulation” pattern.
What is $2.1? Slightly below the midpoint of the pump range, beneath a dense zone of trapped holders. Above, there are holders everywhere waiting to break even; below, support has not yet been genuinely tested.
The bulls and bears are battling—take a look yourself
On one side:
Trump’s pro-crypto policies are still moving forward, including the CLARITY Act
BTC has climbed above $80k, and overall market sentiment is relatively warm
Meme coins have the greatest elasticity; another tweet could drive a 50% rally
On the other side:
The rumor has collapsed + the son’s denial has damaged the narrative
990k wallets are down $3.8 billion, creating an enormous overhang of trapped holders
Insiders unlock $1.9 million worth of tokens daily, with another major unlock in September
Warren and other lawmakers are demanding that the SEC investigate, creating regulatory risk
It has risen 160% from $1.4 to $3.6, and the pullback is far from sufficient
Overhead resistance: $2.24–$2.32 → $2.70 → $3.20–$3.68
Downside support: $2.00 → $1.87 → $1.50–$1.40
Trading strategy
Short-term traders:
Use a small position to hunt for an oversold rebound at $2.0–$2.1, set a stop-loss at $1.96, and reduce the position at the $2.24–$2.32 target zone. If the rebound stalls around $2.3, consider a short position, targeting $2.0 with a stop-loss at $2.42.
Swing traders:
Wait for a high-volume candle with a long lower wick to appear on a pullback to $1.50–$1.70 before considering a long position. Opening a long at $2.1 now offers a very poor risk/reward ratio—the area above is filled with trapped holders, while there is still 20–30% downside.
Long-term believers:
This is not a value coin; there is no such thing as “dollar-cost averaging” here. Wait until a major unlock triggers panic selling before considering it.
TRUMP is not an asset that “rises whenever Trump wins”—
It launched in January 2025, reached a premium of $75, and then fell to just over $1. History has already proven that the brand can drive a pump, while unlocks and selling can completely crush the market.
Insiders are unlocking tokens every day; rebounds provide their liquidity, not an opportunity for you to express your faith.
What is your TRUMP cost basis?
At $2.1, would you dare buy the dip?#BTC回调至79000美元 #老用户1BTC回归礼 #英伟达财报周 $BTC $ETH $TRUMP

Dear Chats!
I’ve been looking at DataVLT and I actually like the direction they’re taking
AI keeps getting smarter, but it still needs something very basic
Real data from the real world
And that data needs to be traceable and verifiable if you want AI, Web3 or DePIN systems to actually use it
That’s where DataVLT comes in
Their new product stack is now live
DVLT ONE connects directly with real-world devices and captures data at the source
DataVLT DID then gives each
contribution a verifiable identity and data trail
So the data coming from the physical world can be connected, verified and used by intelligent systems
I think this is one of those
infrastructure pieces that can be easy to overlook when everyone is talking about AI applications
Because good AI still needs good data underneath it
And Web3, DePIN and IoT all need the same thing
Reliable data you can actually trust
The new DataVLT website is live now
Go take a look if you’re interested in AI X Web3 X IoT X Real-World Data
If you are interested in learning more about it, please follow @DataVLTAI for future updates

I can’t understand it anymore. Some KOLs got liquidated and are still analyzing the market, while people below are frantically asking to copy their trades. I really can’t understand it. Other people have already been liquidated, and you still trust them? Then the liquidated KOLs assume their previous methods were somewhat flawed, review and reflect on what happened, and think they’ve already identified the problems so they can continue making good trades. However, the cruel truth is that what you most need to review and change is this: quit futures trading and reject short-term trading.

Here’s a little-known fact
Even among people at the bottom of society, those in rural areas have a higher chance of knowing big shots than those in cities.
Among the classmates and relatives of people at the bottom in rural areas, some have gotten rich, and they can still run into them from time to time—sometimes even have a meal, drink, and shoot the breeze together.
People at the bottom in cities find it difficult to see the exceptional people among their former classmates and relatives again.

SNDK at $1450—have you been shaken out?
First, the surface picture: a surge followed by a pullback, with retail investors panicking and shouting “double top.”
Q4 earnings on August 5 came in well above expectations, and the stock violently rebounded to 1828 before falling continuously to 1416, a decline of more than 22%. The candlestick chart tells you: 1416 was the August 24 low, 1400-1450 is the decision-making range, and the 20-day moving average at 1380-1410 is sitting right along the lower boundary. A direction is about to be chosen—make sure you are on the right side.
First: earnings were explosive, but the stock did not rise—because they were “not crazy enough.”
Q4 revenue surged 372% year over year to $8.97 billion, gross margin was 84.6%, and EPS was $39.25, all beating expectations. But Q1 guidance was $10.3-$10.8 billion, while the most optimistic in the market expected more than $10.8 billion. That tiny shortfall triggered an immediate valuation sell-off after hours.
Does that sound familiar? NVIDIA has gone through the same script, AMD has gone through it, and now it is SNDK’s turn.
Second: the business model has already changed, but you are still looking at it through an old lens.
Sandisk used to be a cyclical NAND flash stock—making money when prices rose and losing money when they fell. What is Sandisk now?
An AI inference storage provider. Multi-year, floor-price long-term contracts + prepayments + $93.9 billion in guaranteed revenue, an 80%+ gross-margin target locked in for FY28-FY30, and all excess cash returned to shareholders.
It used to live off spot prices; now it relies on profits locked in by long-term contracts.
It used to be a cyclical stock; now it is a growth stock.
It used to trade at 10x PE; now even 30x is considered cheap.
The market is still pricing it based on the “memory cycle,” but its business model has already become an “AI infrastructure rent collector.”
Third: the technical picture has reached a level that must be taken seriously.
1416 was the August 24 low, while 1400 is both a psychological level and support from the 20-day moving average. If it holds, a double-bottom structure is confirmed, with rebound targets at 1500-1570-1600. If it breaks, the next stop is 1330-1350, or even 1200.
From the June ATH of 2354 to 1200 at the end of July, it fell 50%; it then rebounded to 1828, retracing 61.8%, before pulling back to 1400+. This is a textbook “second retest of the lows after a bear-market rebound” structure.
You can judge the bull-bear battle yourself.
On one side:
Q4 revenue surged 372%, gross margin was 84.6%—historic-level data.
A $93.9 billion guaranteed-revenue contract, with most of FY27/FY28 locked in.
$6 billion+ in buybacks, with $15.5 billion in remaining authorization, no debt + net cash.
Analyst price targets of 2000-2250, implying 40-55% upside from current levels.
1400-1416 is supported by the 20-day moving average and a psychological level, making for strong technical support.
On the other side:
Guidance was slightly below the most optimistic expectations—“good, but not good enough” for two consecutive quarters.
The stock has already broken below the 50-day moving average, forming a short-term bearish structure.
Three bombs—NVDA earnings, PCE, and Jackson Hole—are concentrated between today and Friday.
If 1400 fails to hold, the next stop is 1330, followed by 1200.
Trading strategy
Short-term traders:
Try small long positions at 1416-1450, with a stop-loss below 1380 and targets at 1500-1570. If the rebound reaches 1500-1550 but fails to break through, consider shorting, targeting 1420-1400.
Swing traders:
Wait for confirmation of the direction. Go long above 1500, targeting 1570-1600; a breakout opens the way to 1680+. Go short below 1400, targeting 1330-1350.
Long-term believers:
Build positions in batches below 1400, based on the logic of AI storage long-term contracts + a business-model upgrade. Target: above 1600, look for 1800+; in 2027, look for 2000-2250. But note—the three days from August 26 to 29 (NVDA earnings + PCE + Jackson Hole) will be extremely volatile. Reducing positions and waiting on the sidelines is recommended; add back after the storm passes.
For the medium term, the bet is on the AI storage long-term contracts, but 1450 looks more like the dividing line between “a continuation point for the rebound or a second retest of the lows,” rather than a level that is “so cheap you must go all in.”
The August 5 earnings report convinced the market that SNDK is not a cyclical stock, but the August 17-24 pullback has made the market start doubting again. If 1416 holds, it is a golden pit; if it fails, it marks the beginning of a deeper correction.
SNDK now looks like NVIDIA in May 2023—
Explosive earnings but guidance “slightly below the most optimistic expectations,” followed by an after-hours sell-off, and then…
Do you remember what happened afterward?
You are afraid of a pullback; institutions are afraid of missing out.
What is your cost basis for SNDK?
Do you think 1416 will hold? #BTC回调至79000美元 #老用户1BTC回归礼 #英伟达财报周 $BTC $SNDK $SKHY

Huge bullish news!!! Elon Musk $SPCX caught aliens in space!

Gold at $4,630—would you dare chase it?
Look at the surface first: the charge toward 4,700 failed, and retail investors are panicking and shouting “double top.”
Yesterday, it touched 4,696–4,700, a three-month high, then pulled back to the current 4,630. PCE data will be released tonight, with the Jackson Hole meeting taking center stage on Friday. The market is holding its breath. 4,600–4,615 is a strong support zone, the moving averages are in bullish alignment, and the broader trend remains intact. Don’t get shaken out.
First: 4,700 failed to break three times, but 4,600 held three times.
The first push toward 4,700 was slammed back to 4,500. The second push toward 4,700 was slammed back to 4,550. The third push toward 4,700—now it has pulled back to 4,630 without even touching 4,600.
That is called “consolidation at the highs,” not a “trend reversal.”
At the same level, each sell-off makes a higher low. This shows that someone has been quietly accumulating below 4,600.
Second: central banks are buying while you are afraid.
China’s central bank increased its gold holdings by another 20 tons in July, extending its streak of purchases to more than 20 consecutive months. Poland, Turkey, India—the world’s central banks are all buying.
The Treasury is expanding its buybacks of long-term bonds, while expectations for real interest rates are loosening. What do U.S. Treasury buybacks mean? Confidence in the dollar is being discounted, and gold is once again becoming the “ultimate reserve.”
Central banks are not buying gold to trade short term; they are treating it as a “ballast.”
You fear the Fed raising rates; central banks fear the dollar collapsing.
Official institutions think gold at 4,600 is absurdly cheap.
Third: tonight’s PCE is the real battlefield, but 95% of people are preparing the wrong way.
The market expects core PCE YoY to remain at 3.2–3.3%, with MoM around 0.2%.
Weaker than expected → the dollar falls, gold surges to 4,665–4,700, or even makes a new high
Stronger than expected → September rate-hike expectations return, and gold plunges to 4,555 or even lower
Neutral data → range-bound trading between 4,615 and 4,675, waiting for Friday’s speech to set the direction
PCE is only the warm-up; Friday’s Jackson Hole speech is the real “dragon-slaying sword.”
Resistance above: 4,665–4,675 → 4,690–4,700 → 4,720
Support below: 4,615–4,620 → 4,600–4,585 → 4,555
Trading strategy
General principle: reduce your position before the PCE release; do not go all in. Wait for the closing level after the data before taking action.
Scenario A: Soft PCE + 4,615–4,620 holds
Scale into longs on a pullback to 4,618–4,632, targeting 4,665 → 4,690–4,700, with a stop-loss at 4,588–4,595
. This is the most likely direction.
Scenario B: Hot PCE + a high-volume break below 4,600
Reduce longs or try shorts on a rebound to 4,635–4,655, targeting 4,585 → 4,555, with a stop-loss above 4,665. Don’t try to hold out—if the data is bearish, accept it.
Scenario C: Neutral data, range-bound trading between 4,615 and 4,675
Sell high and buy low, with the position capped at 30%. Wait for Friday’s speech before deciding on a trend trade.
Position rules:
Keep leverage within 5–10x; slippage is high on gold data days.
Funding rates are close to 0, with longs and shorts balanced, but volatility will explode once the data is released.
Risk no more than 2% of total capital on any single trade.
Gold now looks like Bitcoin in 2023—
99% of people thought it had “risen too much,” but central banks and institutions kept buying, and the price kept rising.
The day 4,700 breaks, you will realize:
It wasn’t that gold was weak—it was that you cut your losses before the data every time.
For those looking to copy the trade, look here
Don’t chase longs at 4,630; wait for the data.
Soft PCE + 4,600 holds → go long toward 4,700
Strong PCE + break below 4,600 → short toward 4,555
Reduce your position before Friday’s speech; don’t bet on the direction.
What is your gold cost basis?
Tonight’s PCE—are you betting long or short?#BTC回调至79000美元 #老用户1BTC回归礼 #英伟达财报周 $BTC $ETH $XAU

If OKX gave me 1 million USDT, I think I’d withdraw some because tomorrow is Crazy Thursday, and I want to have KFC. I’d use the rest to buy OKB, so I could have KFC every Thursday from now on.✌️✌️

$83 HYPE—would you chase it?
First, the surface: skyrocketing all the way from $50 to $83, with retail traders FOMOing and shouting, “$100 is not a dream.”
It has surged 60%+ over the past month, with a single-day impulse move of 11%-25% after Trump singled it out. There is no historical overhead supply in the price-discovery zone, and the candlesticks look textbook-perfect. The weekly/daily charts show bullish alignment, with MA5/10/20/50/100/200 all below the price. The trend remains intact, but the short-term entry point is extremely poor.
First: Trump’s endorsement has opened the door to a CFTC-compliant route.
Around August 19, Trump publicly stated that the CFTC was bringing Hyperliquid into the U.S. in a “fully compliant and legal” manner. Once the news broke, HYPE jumped directly from $50-$60 to above $80, with a single-day impulse move of 11%-25%.
Hyperliquid is transitioning from a “wild-west crypto project” into a “U.S.-compliant perpetuals gateway.” The market is not buying an already-issued license, but rather the option that “U.S. institutional capital is about to pour in.”
Trump can shout once and send it up 30%, but you cannot expect him to shout every day.
Second: AQAv2 buybacks have begun, but the unlock bomb is here too.
Starting August 26, AQAv2 officially began accruing interest and directing it toward buybacks and burns, which is expected to convert a large portion of the returns from its substantial USDC reserves into HYPE buybacks. In addition, the protocol itself handles several billion dollars in daily volume, its fee revenue crushes that of leading public chains, and 99% of fees go toward buybacks and burns—the three buying channels combined make the fundamentals genuinely strong.
But on August 29, 14.18 million HYPE will be unlocked, worth around $1.2 billion.
Historical post-unlock performance: -7%, +1%, -14%. The moves have been mixed, but it has never “ignored the unlock and surged straight up.”
Third: PCE and GDP are due today, followed by Jackson Hole on Friday.
Today (August 26), the U.S. July PCE (the inflation indicator most closely watched by the Federal Reserve) and the revised second-quarter GDP figure are due, while new Fed Chair Warsh will deliver his first keynote speech on Friday. The broader macro backdrop is that BTC, having just rebounded from its weekly chart, is seeing profit-taking at $78,000-$81k, while the dollar, Treasuries, and tariffs are all creating noise.
HYPE has outperformed the broader market this week, but its beta is not low. Cooler-than-expected PCE + dovish Warsh → surge to $84-$87; hotter-than-expected PCE + hawkish Warsh → first dump to $80, or $77-$78 in a deeper move.
The bulls and bears face off—you decide
On one side:
Trump singled out CFTC compliance for entry into the U.S., opening expectations for an institutional channel
AQAv2 buybacks have begun + protocol fee buybacks, creating three buying channels
Weekly/daily bullish alignment, with the trend still intact
After breaking above the previous high of $75-$77, there is no historical overhead supply above
On the other side:
$83 is already the all-time high, with short-term overbought conditions and crowded momentum chasing
The $1.2 billion HYPE unlock bomb on August 29, with three days left on the countdown
PCE + GDP today and Jackson Hole on Friday mean extremely high macro uncertainty
Funding rates are not extreme, but OI is not low; around $83, both longs and shorts getting wiped out is the norm
Resistance above: $83.5-$84 (breakout confirmation) → $85-$87.5 → $90 → $97-$100
Support below: $81-$82 (ultra-short-term defense) → $78.5-$80 (first re-entry zone) → $75-$77 (last line of defense for trend longs) → $73
Trading strategy
Conservative traders (recommended):
If a pullback to $78.8-$80.2 stabilizes (4H volume expands as selling halts), cautiously open a small long position
Chase the breakout only after a daily close holds above $83.8, with a stop-loss at $81.5
Take profits in batches at: $85.5/$87.5/$90. Reduce the position if it breaks below $77; exit if it breaks below $75.
Short-term traders:
At $83, only trade small swing moves by selling high and buying low, within the $80-$83.8 range. Reduce the position and take profit at $83.3-$83.8, then buy back at $79.8-$80.5.
What to do with existing positions:
Cost below $70: Sell 30%-50% around $83 to recover your principal
Cost of $78-$81: Prioritize protecting principal, and move the stop-loss up to $77.5-$78
Cost of $82.5+: Either enforce a strict stop-loss at $80.8 or reduce to a tiny position and wait until after the unlock
Blindly shorting is not recommended. If you must short, treat it only as an overbought correction: open a small short if a rebound to $83.5-$84 fails, target $81/$79.5, with a stop-loss at $84.3.
Trading scenarios for the next 72 hours
Cooler-than-expected PCE + BTC holding above $79K: HYPE first surges to $84-$86, but may still see profit-taking before the unlock
Hotter-than-expected PCE/hawkish Warsh: first dump toward $80, or $77-$78 in a deeper move—that is a better medium-term buying opportunity, not a liquidation point
Flat data + mounting unlock fears: sweep back and forth between $80-$83; whoever chases gets punished
I remain bullish on HYPE in the medium term—both protocol revenue and the regulatory option are still intact.
But $83+unlock+PCE/Jackson Hole stacked together is a textbook case of a “good asset, bad entry point.”
It is not that HYPE is bad—you just chase every time at the top and sell every time at the bottom.
What is your HYPE cost basis?
At $83, would you dare chase it? #BTC回调至79000美元 #老用户1BTC回归礼 #英伟达财报周 $BTC $ETH $HYPE

I really don’t get it anymore. Some KOLs have been liquidated, yet they’re still analyzing the market, while people below keep bombarding them with questions and asking to copy their trades. I really don’t get it—others have already been liquidated, and you still trust them? Then the liquidated KOLs think their previous methods were somewhat misguided, review and reflect on what happened, and feel they’ve already summed up the problems and can continue on their way to make good trades. However, the cruel truth is that what you most need to review and change is this: quit futures trading and reject short-term trading.

The prediction I posted in the community on Saturday has come true for crude oil, which has indeed come down, but gold hasn’t yet. Oil: I tried placing a few short positions but canceled them all. The entry levels were insanely good, around 87. Gold: I also had a short order at 4700, but canceled that too. I was only betting on a small retracement in the first place. At the current oil price, I would have already closed the position. Besides, having to spend time and energy monitoring them every day isn’t really worth it.

ETH at $2,450—are you chasing it?
Look past the surface: a violent rebound, with retail traders shouting, “Back to the top.”
Up 30% in a week, surging from 1900 straight to 2530, marking the strongest weekly gain since May 2025. But the candlesticks tell you this: the daily RSI has reached the overbought zone at 75, the funding rate has turned positive at an annualized 11%, and the long-short ratio is severely skewed—this is not the initial leg of a major rally, but a high-level digestion phase.
First: a weaker dollar + a short squeeze, not a sudden improvement in fundamentals.
On August 19, expectations of U.S. Treasury buybacks weakened the dollar, and shorts were swept away in one wave. Weekly short liquidations surged, with single-day short liquidations exceeding $32.4k—the price was “squeezed upward” by derivatives, while on-chain fees followed later.
The Glamsterdam upgrade is not due until Q4 2026, and the Fusaka scaling narrative has long been priced in. There is no “immediate bullish shock” from the protocol; the reasons for the price increase have already been exhausted.
Second: institutions are buying, but you need to see how they are buying.
BlackRock bought nearly $700 million worth of ETH in a single week, while BitMine bought another 32.4k ETH last week (about $81 million), bringing its holdings close to 5% of the circulating supply, most of which has already been staked.
Institutions are “allocating”; retail traders are “gambling.”
Third: technically, it has reached a “balance point between bulls and bears.”
The 2500-2550 area above is a dense zone of weekly moving averages + this cycle’s high, while 2415-2380 below is the first support.
Daily RSI at 75 is overbought, and the funding rate is 11% annualized—longs are paying the shorts’ wages. This is not an early-trend feature; it is a late-trend feature.
Only a break above 2550 allows talk of 2700-3000; a break below 2380 means watching 2300 or even 2220.
The bull-bear battle—judge for yourself
On one side:
ETF net inflows of nearly $700 million in a single week, led by BlackRock buying
BitMine holdings close to 5% of circulating supply, accumulating with real money
A 32% staking rate, with the circulating supply continuing to be locked up
The weaker dollar + rate-cut expectations are still building momentum mid-term
On the other:
Up 30% in a week, RSI 75 overbought, funding rate turned positive
2500-2550 rejected three times
New catalysts will not arrive until Q4, creating a news vacuum
Macroeconomic uncertainty ahead of the September FOMC
Resistance above: 2470-2485 → 2500-2550 (strong resistance) → 2700 → 2820 → 3000
Support below: 2415-2380 → 2300-2220 → 2000-2085 (major-level defense)
Trading strategies (no fluff)
Plan A: Buy the dip on a pullback
Wait for a lower wick and a high-volume stabilization in the 2380-2410 range before entering; stop-loss if the daily close falls below 2300, targets 2470-2500 → 2530-2550. Reduce the position at 2530—don’t get greedy.
Plan B: Chase the breakout long
Only chase after a close above 2550 on the 4H or higher timeframe holds, followed by a pullback that holds above 2500; targets 2700 → 2820 → 3000.
Plan C: Short at highs
A spike to 2470-2530 followed by a pullback can be shorted with a small position; stop-loss above 2560, targets 2450 → 2415 → 2380.
Position risk management:
Risk per trade no more than 1%-2% of total capital
A position opened by chasing at the current price should not exceed 30% of a normal position
With the funding rate at 11% annualized, account for the cost of holding long positions overnight
Reduce leverage before the September 15-16 FOMC
What does ETH look like now?
Like BTC in September 2021—
After rising 30% in a week, everyone shouted “100k,” but it took two months of sideways trading before the real breakout.
2450 is not the place to go all in; it is an exam room.
The test is whether you can keep your hands under control.
Waiting for a pullback, waiting for a breakout, waiting for certainty—is 10,000 times more important than betting on direction at this level.
Are you long or short at 2450?
How are you planning to trade this ETH move?#BTC回调至79000美元 #老用户1BTC回归礼 #英伟达财报周 $BTC $ETH $SOL