Orbit Post Sitemap
In June this year, mediated by Qatar and Pakistan, the US and Iran signed a memorandum of understanding containing 14 articles. However, on July 8, Trump directly announced the end of the ceasefire and the resumption of bombing, tearing up all 14 articles at will.
Now the U.S. has again proposed a pause in bombing and "leaving room for diplomatic negotiations," which many people believe outright, even betting 75% of the market on a ceasefire. Yet even Iran's party has publicly stated that they "doubt the U.S.'s intentions." Why are outsiders more certain than the two parties in the conflict? My judgment is clear: the probability of a formal ceasefire agreement reached before the end of August is far below the market price of 75%.
This current rebound is purely a sentiment correction, not a fundamental reversal. Oil prices have fallen for several days, inflation concerns have temporarily eased, risk assets are just catching their breath, US Treasury yields are still stuck at a high of 4.63%, and the Federal Reserve is scheduled for a meeting on Thursday, so the high interest rate environment hasn't changed at all.
Bitcoin has reached $65,000. This level is both a psychological barrier and a key technical battle zone. Bulls hold firm while bears wait for an opportunity. If even one more variable occurs in the Middle East—Trump tweeting or Iran making tough statements—$65,000 could become a stage ceiling at any moment.
The market pattern has long been clear: if good news rises for one day, bad news falls for three; when oil prices fall, BTC rises; when oil prices rise, BTC falls. Most people are always chasing gains, always taking over, always waiting for a break-even. #美联储周四凌晨公布利率决议 $BTC Here are a few new points:
1. The market is essentially bottoming out, and it may take several weeks to build a bottom.
2. The absolute core theme remains AI computing power; other sectors cannot take it away.
3. Overseas CSP investment continues to increase, free cash flow has just turned negative, and there is still significant debt space—much safer than during the internet bubble in 2000. But don't expect Capex to double every year—how could that be possible? In other words, overall growth rate is only a matter of time, but that's not the main point—the key is to focus on segments where Capex growth is greater than the rate of Capex.
4. Optical communications: the growth rate in the next two to three years will exceed Capex's growth rate. Next is the PCB.
5. Overseas storage and domestic storage are not the same thing. I don't see many people understanding this. Overseas it's HBM, but we are not. So you can see overseas CSPs signing long-term contracts with HBM manufacturers, which is rare on our side. Especially those module factories, which are very dangerous.
Regarding domestic storage chains, it is clear that equipment manufacturers have the best logic.
5. Domestic computing power chains also have opportunities, but you must be selective. For example, server switches—I think they're short logic, very short, possibly even shorter than storage module manufacturers. For those with truly medium- to long-term logic, domestic GPUs are definitely the core. Computing power leasing requires even more careful selection; only a few companies can truly make it through.
The night has passed, and dawn is about to arrive. #长鑫科技上市, global storage competition adds variables to $ETH 📺 Gold purchase search volume plummets 80% from its peak: Is the real trading window only after the crowd disperses? Judging from Google search popularity, this round of gold rally is undergoing a very typical process: early on driven by central bank and Asian buying, then in the mid-to-late stage, retail investors concentrated in the market. Now, although market enthusiasm is rapidly fading, gold prices have not collapsed in tandem.
From 2021 to mid-2025, gold rose from $1,800 per ounce to $3,300, but Google search popularity for "buying gold" has barely changed. Meanwhile, central banks and Asian buyers continued to absorb physical gold, while speculative funds and ordinary investors were largely absent, indicating that the early stages of the previous bull market were mainly driven by long-term allocation demand rather than public sentiment. Smart money is still doing what they do best: quietly buying in before the crowd is paying attention.
The real emotional frenzy began in August 2025, with searches for "buying gold" surging rapidly and reaching about eight times the previous level by mid-February 2026. At that time, gold prices were approaching a historic high of $5,600 per ounce, and searches for "buy gold" were nearly eight times those for "sell gold," leaving the market almost entirely with one-way bullish and rally demand. Gold prices have risen about twice from the 2021 low, but public attention has increased eightfold, indicating that sentiment has expanded far faster than fundamentals, which is closer to a crowded trade than rational allocation.
Currently, the search activity for "buying gold" has dropped nearly 80% from the February peak, basically returning to pre-acceleration levels, but gold prices still hover around $4,100 per ounce, only about 20% below the historical high, and still roughly twice the starting point of this bull market.
In other words, market sentiment and speculative bubbles have clearly cooled, but gold prices have not fallen along with the crowd. This is not the typical burst of a bubble; rather, it indicates that central bank gold purchases, monetary credit concerns, and long-term allocation demand continue to support gold prices.
And when everyone stops mentioning gold and prices refuse to continue falling, that's when we should pay close attention.
$XAU I've been watching $BTC's trend all week, so I'm writing some observations.
Many people trade only based on price, not volume-price relationships. A pullback with shrinking volume is a good thing; a drop with increased volume is the signal to run.
Key positions: The upper part is the recent high, the lower is the previous low. Whichever side breaks through first, that's where you go.
Set stop-losses and don't get carried away. Only by staying alive can you have a chance.
BTC / #BTC$WDC is currently trading at $534.74 on OKX following a strong bounce from $432.
The price is consolidating, with key resistance at $545.00 and support at $525.14—a breakout above resistance could spark the next bullish rally.
#DailyOrbit @OKX中文 Trump has three choices: hit, suppress, or withdraw. Which path do you think Trump will ultimately choose?
When even the president himself doesn't know what to do next, the market is the biggest casino. The New York Times revealed that Trump is being put on the stake over the Iran issue—military escalation, economic strangulation, or a dignified retreat? Internal disputes are in chaos: sanctions haven't crushed Iran, withdrawal is for fear of trouble in Hormuz, and even more for fear of getting caught up in it.
Geopolitical uncertainty premium. If oil prices soar and inflation expectations rise, it will be even harder for the Federal Reserve to manage; But if it really comes to war, safe-haven funds will briefly rush into the big market, but remember—the early stages of the war rise quickly, and the mid-term drops are also steep.
Retail investors shouldn't gamble on national fortune with candlestick charts. My view: Old Te is very likely to choose "fight while negotiating," but the market will be slapped back and forth. At times like this, watching the show is better than acting. #美军暂停对伊空袭, international oil prices opened sharply #交易之声: Your experience deserves to be heard #美国禁止开源AI的预期大幅回落
Is the US going to ban open-source AI? After all this fuss, it was decided who made the money
Recently, rumors spread that the U.S. plans to completely ban open-source AI, causing widespread anxiety within the industry. But after a week, the probability of the ban dropped from 60% to 19%, and expectations were completely dashed.
There's no sudden policy shift; essentially, it's two groups of American tech companies fighting, with the one with more money and a longer supply chain winning.
One group is OpenAI and Anthropic, who make a living by selling APIs, naturally hoping that all of open source will die out, wanting to monopolize pricing and constantly complain about "open source isn't safe" in the White House;
The other group is Nvidia, Microsoft, and Google—the shovel-selling companies. Their chips, cloud services, and developer ecosystems all rely on open source, and banning open source would cut off their revenue streams. They directly brought in 50 giants to jointly submit the petition, along with over 200 startups siding with them, making the situation fully charged.
The result is realistic: regulators don't dare offend anyone, but since shovel sellers create more jobs and pay more taxes, it's naturally up to them to call the shots.
Finally, they implement tiered regulation and just focus on the top-tier cutting-edge models, playing as usual open-source as they wish.
In the end, on the surface, it's all about safety and risk, but behind the scenes, interests always determine policy direction.BTC is now at 62,000. Let me share how I see it now
On-chain data has recently shown 3 signals
I've been watching the market for six years and have seen similar scenes 4-5 times.
Realized losses of $3.5 billion. Losses from selling chips are absorbed, and the market clearance is nearing its end.
The miner selling index is at its high. The cash cost of the S19 series mining machines is 60,000, and the current price is close to cash cost.
The exchange balance dropped to 1.98 million. In previous major drops, retail investors have moved coins into their wallets; 1.98 million is the lowest since 2018.
Patience and discipline are more important than predictions.
No one knows the bottom, so don't worry.
📌 Put this signal back into the market structure
Price, trading volume, and on-chain data should be viewed together. If the price falls but long-term holders do not reduce their positions simultaneously, it usually indicates that the chips are being reallocated; If the exchange experiences both net inflow and rising leverage, you should first treat it as a risk signal rather than rushing to guess the bottom.
🧭 How will I track them?
First, see if the key price level can be pulled back for two consecutive days. Second, check whether spot trading volume keeps up, rather than just looking at contract volatility. Third, look at the direction of the incoming and outgoing transfers of large denomination addresses. When these three layers of signals do not resonate, I reduce my position and wait for the market to give me an answer.
⚠️ Risk reminders
On-chain addresses do not represent a person's full intent; exchange wallets may also be merely internal scheduling. No single data point can directly convert into buy or sell orders; position size and stop-loss discipline are more important than prediction.
🎯 The final execution framework
First, use small positions to verify judgment, then determine whether to add positions based on price and volume; If the basic assumption is broken, exit without arguing with the market. The purpose of this approach is not to guess correctly every time, but to keep errors within a manageable range.
I'll break this topic down into three layers. The first layer is data that can be directly observed. First, record values, time, and direction, avoiding jumping to conclusions based on just one screenshot; The second layer is how the market reacts: data improves but prices remain unchanged, and weakening data but prices still rise—the meaning is completely different; The third layer is your own operations: first write down your maximum tolerable loss, then decide whether to adjust your position. This sequence may seem slow, but it helps reduce being carried away by a single headline.
For me, price, spot volume, and on-chain tokens should be compared on the same table. Each update only changes the parts with new evidence; a single change in number cannot overturn the entire judgment. If the three observation directions contradict each other, I would downgrade the conclusion to 'waiting for confirmation' rather than forcing a bullish or bearish story. The most easily overlooked cost in the market is determining it too early and then refusing to admit that the assumption has failed.
In practice, I first use observation positions to test and wait until at least two of the trading volume, price, and fundamentals are aligned in the same direction, then consider increasing exposure; If volatility increases or liquidity thins, reduce your position first. Any backtesting, historical cases, or KOL perspectives can only be used to establish hypotheses and cannot replace current risk checks. This article is my research notes, not buy or sell orders that guarantee profits.
In my next update, I will re-examine four things: whether the message is still valid, whether the price reaction has been confirmed, whether liquidity is sufficient to execute, and whether the original risk assumptions have been broken. If it's just a rise in social media buzz without seeing trading volume or capital support, I treat it as a signal to watch; If the data direction changes, the original script will be updated accordingly, rather than holding it for the sake of saving face.
The advantage of this approach is that it separates "perception" from "action." Opinions can retain multiple possibilities, but actions must have clear triggering conditions. For short-term trading, I set a time limit; For medium- to long-term allocations, I will check fundamentals and capital costs. No matter the final outcome, record the reasons for entry, exit, and actual slippage, so that next time you'll have real material for improvement.
If sources conflict with each other, I will mark the conflict first and wait for confirmation in the original announcement or the next time, rather than using social media sentiment as evidence. This also means that sometimes the best strategy is to wait without a position, because not trading itself is also a way to manage uncertainty.Guys, just look at these numbers—this car is damn heavy!
KAITO's long-short ratio jumped straight to 313%, all waiting to get rich in the car? Looking at Smart Money's data, the bulls are lying low at 0.72, and now they're floating with nearly 18 million U in profit! 18 million dollars? Do you expect the dog farm to sing praises and talk about vision here?
Dog farms have never been driven by charity; now they're just holding on at high prices and letting retail investors take over and sell off. The bears have been mostly washed out, and the upcoming script can be guessed with eyes closed—definitely a reverse sell-off. With so many profit-taking positions hanging above, as long as the main players lead the sell-off, everyone below is trampled. Going long at this position is pure big frustration. Old Liu, I'm just emptying my head and waiting for a big waterfall!
---
Market analysis
Let's start with KAITO — today this stock is indeed strong, with spot trading up 18.22%, reaching around $1.1931, and hitting a 24-hour high of $1.2318. But if you look closely, the Binance large account long-short ratio reached 2.74. What does that mean? The large holder's long position is nearly three times that of the short position! Such extreme position structures have historically often signaled imminent market changes. More importantly, this rebound rose from the low of 0.9553, with profit-taking piling up like a mountain. The psychological barrier of 1.20 was pressing down like a fortress. The buying depth is decent at 1.10, but the funding rate is only 0.0028%—what does that mean? Bullish leverage hasn't been sufficiently leveraged; the main players have no intention of pushing the market at this level.
Now let's look at SHIB — it rose 36% over the weekend, dropping from 0.0000042 to 0.0000058, with daily trading volume peaking at $380 million. But brothers, calm down, what is this rally relying on? FOMO sentiment among Korean retail investors is as follows: just the SHIB/KRW trading pair accounts for one-tenth of the global total. Another whale wallet that had been dormant for six months suddenly activated, spending $125,000 to buy 30 billion SHIB. Sounds pretty intimidating, right? But let me tell you, the daily RSI has reached extremely high levels, and the Bollinger Bands %B indicator has surged to 1.06—this is textbook overbought status. On the analyst side, CoinCodex's year-end target price is only 0.00003422, which is an 18% drop from now. Moreover, this rebound lacks even a decent fundamental catalyst.
DOGE is even more outrageous—top traders' long-short ratio is 3.54%, and 78% of so-called "smart funds" are all long. Retail investors accounted for 73.6% of long positions. This isn't institutional confidence—it's crowded trading! Open interest is still falling, with a buy-sell ratio of 0.985, with sellers pushing buyers to buy. The daily RSI is almost at 100, and wherever you put this data, it's a signal of a pullback. DOGE is currently at 0.07277, with support below 0.0725. If it breaks through, it could move straight to 0.068-0.065.
On the market side — Early this morning, news of a US-Iran ceasefire was released, Bitcoin returned to 65,000, and Ethereum rose over 3%. But the Fear and Greed Index is only 26, still in the fear range. Moreover, tech giants have just finished earnings season, and Tesla still holds 11,509 BTC, making it uncertain whether Alphabet's cloud business can sustain its growth. How long the market can maintain this momentum is hard to say.
---
Trading direction and trend strategies
KAITO — I choose to go short. Open a position in the $1.16-1.19 range, set a stop loss above $1.23 (previous high resistance), first target $1.05, second target $0.95. The logic is simple: the long-short ratio is 313%, while the long-short ratio for major players is 2.74. With such extreme data, if the main players don't harvest the bulls, who else will they do? The resistance at 1.20 is as tough as a solid plate; after three pullbacks, it hasn't broken below 1.166, indicating there is indeed buying below, but the upward momentum has already been exhausted.
SHIB — Mainly wait-and-see, supplemented by short selling on rebounds. The weekend's 36% surge has already exhausted sentiment. The daily RSI is overbought, and the upper Bollinger Band is being blocked. Chasing long positions at this level is the way to buy the position. If it rebounds to the 0.0000054-0.0000055 range, you can take a light position and try shorting, with a stop loss at 0.0000059 (resistance at the 200-day moving average), and a retracement target at 0.0000048-0.0000045.
DOGE — waiting for opportunities after a pullback. No rush to act now; wait for it to break below 0.0725 to confirm the direction. If it really falls to the 0.066-0.069 range, it would be a good entry point for medium-term long positions. But the premise is—wait for support to be confirmed and trading volume to shrink, and don't rush to bottom-fish.
---
Trading insights
Brothers, after so many years of trading, I've come to realize one truth: the market always rewards patience and punishes FOMO.
When SHIB rose 36% over the weekend, how many people slapped their thighs and said, "If only I had known, I would've gotten on board"? When KAITO rose from 0.96 to 1.19, how many people chased at the high level and are now stuck in the gap? Behind every big bullish candlestick you see, there is a bunch of people standing guard at the summit.
The dog farm's trick hasn't changed for years: accumulate shares → drive prices up→ sell off→ dump the price. Currently, both KAITO and SHIB are in a transitional phase from "selling off" to "dumping," with shockingly high long-short ratios, overbought RSI, and profit-taking positions—these signals are right there, and if you insist on charging in, that's not bravery, you're just handing yourself over.
Also, don't go against the data. The long-short ratio is 313%, the long-short ratio for big players is 2.74, and the RSI is close to 100—none of these are meant to scare people, but to save lives. I'd rather miss a rebound than stand guard on the mountaintop waiting for a breakthrough.
Remember Lao Gao's words: Better to earn less than to lose big. As long as the capital remains, opportunities will always be there; If your principal is gone, even if the dog farm pulls you up, it has nothing to do with you.
$KAITO $SHIB $DOGE
#财报观察员: Who can truly understand the real answer sheet from Google and Tesla this time?
$SHIB $DOGE $KAITO
#长鑫科技上市, global storage competition adds new variables
#财报观察员: Can Microsoft, Meta, and Amazon stabilize the AI narrative?
#美联储周四凌晨公布利率决议 This week could determine your earnings for the second half of the year.
Wednesday is the Federal Reserve, Thursday is Apple and Amazon, plus tariffs taking effect and soaring oil prices, four major events detonating in the same week.
I'll lay out the timeline for you this week.
Wednesday afternoon, the Federal Reserve FOMC decision. Hawkish Chair Warsh meets amid a rebound in inflation; the probability of a rate hike is low but not zero. This is the master switch for market sentiment.
After the market closes on Thursday, Apple and Amazon earnings reports. The final battle of the seven giants; the market will use them to judge whether the whole AI capital expenditure story can continue.
In the background, there are two other factors. Global tariffs pushed to 15%, increasing inflationary pressure; the Iran situation remains unsettled; oil prices are still high.
Four events packed into one week, the outcome of any one could trigger severe volatility.
In such a week full of uncertainty, I won’t take any aggressive actions, keep my position at a level I can fully withstand, and keep enough cash on hand.
No guessing the Fed, no betting on earnings, no short-term trades.
Because the biggest feature of this week is huge volatility but random direction.
Your chance of guessing the right direction is about the same as flipping a coin, but if you guess wrong and use leverage, the losses are real and tangible.🔥 Why can't most people make big money in a bull market?
Not because they don't know how to buy.
But because—he couldn't hold on.
📉 A 5% drop makes you start questioning your life,
📈 If it rises 20%, you rush to get out.
In the end, watching others double their wealth,
But he only earned a little "pocket money."
A truly great opportunity,
It was never something you could do every day,
It was the result of enduring it.
In this bull market, I stick to only three things:
✅ Hold core assets
✅ Avoid frequent portfolio rotation
✅ Patiently wait for the cycle to be fulfilled
The market keeps making noise,
But wealth always rewards those who are patient.
Remember this phrase:
**You won't make money beyond your understanding, but you will definitely earn money from patience. **
I believe that the real opportunity to change the fate of ordinary people in the coming years still lies in the crypto market.
Time will tell.
🚀🚀🚀
#BTC #ETH #SUI #OKB #Crypto #比特币 #以太坊 #牛市 #Web3 #长期主义📊 Cross-asset quotes | 19:26
EUR/USD 1.1391 (+0.19%) / USD/JPY 163.63 (-0.14%) / USD/RMB 6.7660 (-0.06%)
Volatility clues: EUR/USD changes are more evident; first observe whether this affects dollar liquidity and risk asset sentiment.
Observation perspective: Quote-type content and main account updates are staggered, suitable for supplementing external variables in the crypto market for precious metals, energy, and forex.
Verification point: If these assets diverge from BTC/ETH, prioritize whether risk appetite is being repriced.
For market observation purposes only and does not constitute investment advice.$ZAMA** 🧬
**+7.32% – MACD bullish crossover confirmed. Momentum is BUILDING.**
Price: $0.05901. RSI6 at 58.33 – room to run. MACD: 0.00023 – bullish. SAR at $0.05765 – price above. 24h low at $0.05017 is miles away. Break $0.06041 and we target $0.064+. 🚀
Also watching: WLD🌐, KAITO 🤖, $BEAT 🎵📊 $SAND Market Outlook
This week, $SAND is approaching a critical technical zone.
The first level I'm watching is 1,412. If sellers manage to break below that support, the next downside areas to monitor are around 1,382 and 1,360, where buyers could attempt to stabilize price.
That said, the larger trend may not fully shift until 1,300 is decisively lost.
📉 Key Levels
🔻 Support: 1,382 → 1,360 → 1,300
A confirmed breakdown below 1,300 could open the door to a much deeper decline. Rather than happening in one sharp move, the downside could unfold in stages, creating both short-term rallies and fresh trading opportunities along the way.
📌 Trading Perspective
• Above 1,300, the market still favors short-term range trading rather than a confirmed trend breakdown.
• If 1,300 fails, expect volatility to increase, with relief rallies likely appearing before the next leg lower.
• Positioning matters. Chasing shorts after an extended drop often offers a poor risk-to-reward profile. Waiting for higher-probability entry zones usually provides better trade management.
The focus isn't on predicting every move—it's on reacting to confirmed price action and managing risk with discipline.
#CXMTMemoryIPO #FOMCRateWatch #AIEarningsWatch 📊 Monday Market Outlook: $BTC & $ETH
Last week's roadmap played out largely as expected. After strong relief rallies, $BTC tested the $67K region while $ETH pushed toward $1,960 before both saw a pullback.
Now the market faces a key question:
Is this the beginning of a new uptrend—or simply another relief rally?
For now, I'm waiting for stronger confirmation.
Recent strength has been supported by improving macro sentiment, with easing concerns around geopolitical tensions, oil prices, and inflation. Even so, the broader market structure has not yet confirmed a decisive bullish reversal.
👀 Key factors to watch:
🔹 ETF flows continue to provide insight into institutional participation.
🔹 Momentum has improved, but conviction remains limited.
🔹 $BTC still needs a strong breakout above the $65.5K–$65.8K resistance zone.
Until that area is reclaimed with convincing volume, caution remains the preferred approach.
🟠 $BTC Trade Levels
📍 Entry: ~$65,500 & ~$66,300
🎯 Targets:
• $64,500
• $63,600
• $62,800
🔵 $ETH Trade Levels
📍 Entry: ~$1,960 & ~$1,980
🎯 Targets:
• $1,920
• $1,880
• $1,840
⚠️ Stay disciplined, manage your risk, avoid oversized positions, and let price action confirm the next move before committing to a trade.
#CXMTMemoryIPO #FOMCRateWatch #AIEarningsWatch
$ETH $BTC $SHIB In terms of data, Ethereum remains the absolute king of DeFi. TVL accounts for over 60% of the entire network, stablecoin supply exceeds half, and the number of developers leads by a wide margin. Although L2 diverts transaction volume, it also allows the mainnet to focus more on settlement layer value. In terms of ecosystem depth, other chains cannot catch up in the short term. The sluggish coin price is not due to poor governance. The core issue is that mainnet fees have sharply decreased after the rise of L2, deflation expectations have been dashed, and macro liquidity has tightened. Selling coins is an open operating expense; while the pace is not good, it is fundamentally different from "dumping the stock for cash." As for poking fun at Vitalik's personal life, for a developer who consistently outputs core code and academic papers, it is both unfair and misses the essence of the issue. 2026 will indeed be a turning point for reform. The foundation has launched its largest recent adjustments: restructuring management, focusing on application-layer funding (RWA, stablecoin payments), and enhancing financial transparency. The direction is correct, but the results will take time to prove. The "surprise" of the next bull market will not be a move by Valve or the foundation, but whether Ethereum can turn its technological advantages into real fee capture capabilities. Multi-chain coexistence is already a done out; ETH prices must be driven by actual revenue, not mere narrative premiums. The moat exists, but the new weapons are still not ready. Reform is just starting and worth looking forward to, but don't expect immediate results. #以太坊验证者退出队列已降至零 7.27 Gold
From the current trend, the channel is gradually closing and expanding again, with the point running above the middle band and near the upper band, indicating that bullish strength is accumulating and holding a short-term dominant position.
After forming a golden cross above the zero axis, the MACD indicator continues to release positive momentum. Although the histogram has converged somewhat, it remains in a positive range, indicating that the upward trend has not yet exhausted. The pullback is more of a technical correction than a trend reversal. In terms of candlestick patterns, after a recent consecutive bullish rise, there has been a slight consolidation with some volatility, but the lows are gradually rising, and support levels are solid, reflecting strong buying support. After breaking through the previous consolidation platform, the price did not show significant volume stagnation or prolonged upper shadow suppression; instead, after confirmation of pullback, it stabilized and rose again, indicating a somewhat optimistic market sentiment.
(4079 advances, 4064 supplements, 4044 defensive, watch 4145-4185)
The above is an objective analysis of the market and does not constitute any investment advice. The market is subject to uncertainty. Please make specific trading decisions based on real-time market conditions and your own risk tolerance, and make independent and prudent judgments and bear the corresponding risks.
$ETH $BTC $XAU 下周开盘前需要知道的几件事
周末传来两个好消息
韩美这周敲定了一份规模看着吓人的AI半导体合作框架。名义上9500亿美元，三星、海力士、英伟达、博通都签了字，海力士和英伟达那笔HBM合作占了7500亿。但这个数字水分不小，本质是未来五年才逐步兑现的意向协议，实打实每年能落地的也就千亿出头。这种长约对海力士未必是好事，万一现货价跌破约定价格，反倒要自己承担违约或者压毛利的风险。英伟达则轻松把产能过剩的风险甩给了制造端，稳赚不赔。实打实的利好是此前压在市场心头的韩国养老金减持担忧终于解除了，7月数据显示这笔钱年内头一次转为净买入，还专门加仓了海力士。
美伊这边打了13天之后，双方都停手了。表面看像和解，实际是美军的防空拦截弹快打光了一枚。400多万美元，已经消耗了1200多枚，而伊朗一直用便宜的无人机在跟你耗。停火让油价松了口气，通胀压力暂时缓一缓，但红海那边胡塞武装依旧在骚扰油轮，乌克兰这周还在里海打沉打伤了几艘伊朗船，地缘这条线远没有真正落地。油价短期回调后EMA 20 80 接着看涨。
下周三个真正决定方向的变量
一是美联储决议。利率大概率维持不变，真正要看的是沃什怎么措辞，有没有对9月加息留口子，怎么处理油价反弹带来的通胀反复。感觉会偏鹰。
二是日本央行会议。日元继续弱下去，可能倒逼外资抛美债，间接推高美债收益率、压制美股。10年期美债目前创出一个更高的高点到4.7%，超过5月18号的高点，短期回落到EMA 20 4.58%后接着看涨。周线级别突破了这个4，5年的三角整理，5%可能不会是这个周期的顶点。
三是这周扎堆的巨头财报。微软、Meta、苹果、亚马逊、海力士全都在这几天。市场最怕再来一次谷歌那种剧本(开支猛、现金流跟不上、直接被砸)，上周特斯拉跌了近18%、谷歌跌了近8%，这次谁能扛住是关键。Meta看CapEx指引会不会继续往上调。亚马逊看AWS的开支节奏和AI订单能不能落地。苹果现金流最厚，可能是这波财报季里相对安全的选项，目前也走得最强，日线EMA
20一直没跌破，沿着EMA 20一路回踩涨。海力士和三星看HBM出货和毛利。
大盘技术面：偏弱，几个关键位记一下
标普SPY比高点低了不到3%，上方744、750、752是连续几道阻力，走势明显在走"低点更低、高点也更低"这种偏弱结构，下方支撑看736、730、724。周线级别回调不会低于700。
小盘股(IWM)是三者里最差的，一路阴跌，均线上没建立过一次像样的反弹，日线还出现了四重顶背离，周线目标可能看到260-265附近。
VIX短期倒是出现了个顶背离信号，VIX跌美股反弹，验证这两天有反弹空间，加上美联储会议前的观望情绪，周一周二可能会有一波技术性反弹。但强调一下，只是技术性的，别当成反转，会议之后大概率还得接着调整。
基本面这块，其实还挺硬
标普二季度盈利同比涨了38.8%，远超市场原本预期的24%，85%的公司都超预期，这在历史上是相当高的比例。换句话说，盈利涨得比股价快，等于股票正在变便宜，这对愿意拿长线的人是个不错的窗口。消费端也没崩，运通、Capital One这些公司的财报都显示各个收入层级的花钱意愿依然在，坏账率也压得很低。不过要留个心眼，现在全球股票总市值已经涨到全球GDP的137%，跟2021年meme股疯狂那阵子打平了，历史上这种估值位置很少能一直撑住。
季节性：8、9月历史上是弱的月份，但通常是给年底铺路
历史统计里8、9月经常跌，但只要7月底前涨幅到了7%以上，35次里有32次年底都是正收益，涨幅还大多集中在11、12月。所以就算接下来一两个月走弱，历史规律倾向于把它理解成年底行情前的蓄力期。
总结
周末几个消息给了市场喘口气的理由，加上VIX的顶背离信号，周一周二美联储会议前大概率会有一波反弹，但这更像是情绪性的技术反弹。真正的考验在会议之后，超级财报周的现金流焦虑、加息概率被悄悄上修、加上8、9月历史上偏弱的季节性，三个因素叠在一起，反弹以QQQ为主我看几个位置696，700和704缺口，696和700有机会到，704缺口不一定补。反弹完后看震荡下跌，QQQ最终回调目标不会低于637， EMA 200在648，SMA 200在643。
好在盈利数据摆在那里，只要美联储没有意外动作，这波波动大概率还是中期选举年下半年常见的季节性震荡，长线的判断没必要因此改变。$ALLO** 📈
**-7.15% – RSI6 at 31.43 – EXTREME OVERSOLD. Bounce loading.**
Price: $0.31750. 24h low at $0.31500 is holding by a thread. MACD bearish but divergence is massive – reversal incoming. SAR at $0.32475 is the trigger. Break $0.34986 and we run to $0.37+. 💪
Also watching: RE📊, LAB 🧪, $UB 📈#美联储周四凌晨公布利率决议
In the early hours of July 29, the Federal Reserve will hold a meeting to set interest rates again. This time is completely different from before—there was no prior hint, no reassurance to the market. The new chair, Waller, directly scrapped the decade-old "forward guidance" rule. From now on, each meeting will only consider current data, with no room for any forecasts. In plain terms, this is a blind box with no script; no one knows what's inside, and even professional institutions are in uproar.
The market is now completely divided. On one side, dozens of economists unanimously say "definitely no rate hike," believing that since inflation dropped in June and employment remains stable, there is no need to tighten further; on the other side, futures market traders are betting real money with a 38% chance of a rate hike. They fear oil price rebounds, tariff risks, and new inflation driven by AI, thinking current rates cannot suppress prices. This split between "experts say one thing, the market does another" is rare in recent years and shows how difficult it is to predict this meeting—no one dares to guarantee the outcome.
For ordinary people like us, this meeting is no longer just a macro narrative in the news but a reality tightly linked to mortgages and wealth management. Even if the Fed holds steady this time, Waller's hawkish stance, high oil prices, and AI-driven inflation pressures will keep U.S. Treasury yields high, preventing banks' funding costs from dropping. Mortgage rates won’t ease easily, and monthly payment pressure will be real and heavy. Previously, we hoped interest rates might drop in the second half of the year and mortgages might loosen, but now it seems that wish will most likely be dashed. Don’t hold unrealistic expectations anymore.
Facing such uncertainty, the most reliable approach is not to bet on "whether rates will rise or not," but to learn to live with uncertainty. Don’t blindly trust experts’ consensus, nor be led by market probabilities. Pay close attention to Waller’s press conference wording, oil price trends, and inflation data before September. Keep your wallet tight: hold more cash, avoid high-valuation tech stocks, revisit those so-called "safe" wealth management products, and don’t be fooled by superficial stability.Analyze $BNT /USDT current price $0.2862 and generate a professional trading setup using current market structure.
Include:
• Market Trend (Bullish/Bearish/Neutral)
• Key Support Levels
• Key Resistance Levels
• Optimal Entry Zone
• Take Profit 1, 2, and 3
• Stop Loss
• Risk/Reward Ratio
• RSI Analysis
• Volume Analysis
• Price Action & Market Structure
• Scalping View (intraday)
• Swing Trade View (3–14 days)
• Trade Confidence (%)
• Risk Management Advice
Response Requirements:
• Professional trader style
• Data-driven analysis
• Clear formatting
• Maximum 120 words
• Avoid generic statements
• Focus on high-probability setups#DailyOrbit ⚡️ REMINDER: Zcash’s “Ironwood” upgrade is expected to activate on mainnet tomorrow, July 28. $DOGE** 🐕
**-0.50% – MACD just flipped bullish. Meme season loading.**
Price: $0.07269. RSI6 at 51.95 – neutral, ready to run. MACD: 0.00003 – bullish crossover. SAR at $0.07240 – price above. 24h low at $0.07213 is solid. Break $0.07351 and we run to $0.075+. 🚀
Also watching: SHIB🔥, PEPE 🐸, $BOME 🎨🚨 Micron vs. Kioxia — Is the Memory Cycle Finally Turning?
Micron's latest earnings sparked mixed reactions, but one thing is becoming increasingly clear: the AI memory story is entering a new phase.
HBM remains one of the strongest beneficiaries of the AI boom, with Micron reporting HBM revenue up 60% year over year. At the same time, memory names like Micron and Kioxia are beginning to show signs that the sector could be moving out of its downturn.
But one quarter doesn't confirm a new cycle.
📌 Here's what I'm watching beyond the headlines:
🔹 Can HBM production keep up with demand?
🔹 Are higher gross margins sustainable, or just temporary?
🔹 Will cloud providers continue spending aggressively on AI infrastructure, or shift toward optimization?
Strong demand alone doesn't guarantee every memory company will benefit equally.
🧭 My research framework
Instead of focusing only on revenue growth, I compare:
• Order visibility and factory utilization.
• Product pricing, yields, and capital expenditure efficiency.
• Results across memory producers, equipment suppliers, and hyperscale cloud companies.
If share prices rally while fundamentals fail to improve, I treat it as a short-term trade—not a long-term investment thesis.
⚠️ Key risks
The AI narrative has already pulled a lot of future expectations into today's valuations.
Higher supply, slower enterprise spending, or weaker pricing power could quickly change sentiment.
That's why I separate market opinion from investment decisions.
🎯 My approach
I prefer waiting for at least two consecutive quarters of improving fundamentals before increasing long-term exposure.
Until then:
✅ Follow the data.
✅ Compare evidence across the supply chain.
✅ Scale positions gradually.
✅ Let new information change your view when necessary.
The best investment decisions are built on evidence—not headlines.
#CXMTMemoryIPO #FOMCRateWatch #AIEarningsWatch The CLARITY Act (Digital Asset Market Clarity Act, H.R. 3633) is the first federal crypto market structure law in the United States. The House passed the House of Representatives in July 2025 by a vote of 294 to 134, and the Senate Banking Committee passed it in May 2026 by a vote of 15 to 9. The bill is currently stalled in the full chamber—the last window before the August 7 recess.
What is it for: Divide digital assets into three categories—digital commodities (BTC/ETH, under CFTC), investment contract assets (early tokens/securities, under SEC), and payment stablecoins (separate framework); As the network "matures," tokens can graduate from SEC to CFTC classes. End a decade-long enforcement regulation where "SEC and CFTC compete for territory."
Where the stitch is: 60 votes to prevent a long threshold; Republicans have about 50 votes, requiring 7-10 Democrats. The final stumbling block is the ethical clause—banning the president, congressman, or spouse from issuing coins or profiting from crypto during their term. Trump has agreed, but Democrats complain that law enforcement powers are not independent to the DOJ and that consumer protection is insufficient.
Two outcomes: Passed = exchanges/custodians have clear registration paths, institutional funds enter with legal grounds, and the US seizes global rule discourse; Void = Returned to the "SEC lawsuit" gray area, waiting for the new 2027 Congress to deal with it. Galaxy has already cut its 2026 clearance probability from 60-75% to about 50%.
A final sentence: this is not "crypto legalization," but "crypto finally has a legal basis"—but if there is no movement after two weeks, the industry will have to endure in vague stagnation for another year. #财报观察员：微软Meta亚马逊能稳住AI叙事吗？
The trend is back!!!
Microsoft, Meta, and Amazon are about to release their earnings reports. The performance of these three tech giants may determine whether the next phase of the AI market can continue.
Over the past two years, one of the biggest winners in AI has undoubtedly been NVIDIA, with computing power demand driving a boom in GPUs and data center industries. Now, the market is starting to look for new answers:
With such huge AI investments, can it ultimately translate into profits?
Microsoft, relying on the OpenAI ecosystem, holds a leading position in cloud computing and enterprise AI applications.
Azure growth and Copilot commercialization progress will be key market indicators to watch.
If AI service revenue continues to grow, Microsoft will remain one of the strongest beneficiaries of AI commercialization.
Meta is taking a different path.
The market previously worried about Meta’s huge AI investments, but with large models, recommendation algorithms, and improved advertising efficiency, AI is now driving its core business in reverse.
If AI can boost advertising revenue, Meta’s high investment logic will be revalidated.
Amazon’s key lies in AWS.
Cloud computing is an important foundation for AI commercialization. Whether AWS can accelerate growth again will directly affect market confidence in Amazon’s AI story.
In the short term, the earnings reports of these three companies may cause significant volatility.
If AI capital expenditures continue to increase but revenue realization falls short of expectations, the market may once again question the "AI bubble."
But if cloud business, AI services, and advertising efficiency show clear improvement, the AI theme may continue to attract funding.
For the crypto community, the earnings reports of US AI giants are equally important.
If the AI narrative continues to strengthen and risk appetite rises, the AI+Crypto sector may see a return of capital.
Key focus:
TAO, $FET, and other AI sector assets.
At the same time, $BTC and $ETH, as the core of market liquidity, will also be influenced by the sentiment of US tech stocks.
My view:
AI will not end because of one earnings report, but the market will reselect the true winners.
Future AI competition is not just about models, but about computing power, ecosystem, and commercialization capability.
Microsoft, Meta, and Amazon are proving one thing:
The first phase of AI is storytelling; the second phase is the real competition of profitability.
In this AI wave, those who survive will not be all participants, but the companies that can turn technology into cash flow. Wintermute maintains a short position on all coins except $BTC .📊 Monday Market Outlook: $BTC & $ETH
Last week's roadmap unfolded largely as expected. After strong relief bounces, $BTC approached the $67K region and $ETH climbed toward $1,960 before both retraced.
Now the focus shifts to one question:
Is this the start of a new uptrend—or just another relief rally?
At this stage, I'm still waiting for stronger confirmation.
The recent recovery appears to have been supported by improving macro sentiment as concerns around geopolitical tensions, oil prices, and inflation eased. However, the broader market structure has yet to show a decisive bullish shift.
📉 Key factors I'm watching:
🔹 ETF flows remain an important indicator of institutional participation.
🔹 Momentum has improved, but conviction is still limited.
🔹 $BTC has yet to secure a strong breakout above the $65.5K–$65.8K resistance area.
Until that zone is reclaimed with convincing volume, caution remains warranted.
🔻 $BTC Trade Levels
Entry: ~$65,500 & ~$66,300
Targets: $64,500 → $63,600 → $62,800
🔻 $ETH Trade Levels
Entry: ~$1,960 & ~$1,980
Targets: $1,920 → $1,880 → $1,840
⚠️ Manage risk carefully, avoid oversized positions, and let price action confirm the next move before committing to a trade.
#CXMTMemoryIPO #FOMCRateWatch #AIEarningsWatch $XSPCX Looking Bullish
Buyers are showing strength as momentum improves, keeping the trend biased to the upside.
Entry: 114.20 – 115.00
TP1: 118.00
TP2: 121.00
TP3: 125.00
SL: 111.80
Analysis:
Positive price action and steady accumulation suggest room for additional gains if resistance is cleared.
Let's trade $XSPCX
#OKXOrbitTopics 沙特卖油用人民币结算，转头兑换金条运走，会掏空我国储备吗？
中东产油国把石油卖给中国，收到人民币后，转身买走一批金条运回国内。
中国拿到的是烧完就没了的原油，对方带走的却是真金白银，长此以往，中国的黄金储备会不会被搬空？
这笔账看似简单，实际上算错了对象。
产油国拿人民币购买黄金，买的是市场上的黄金，不是拿着人民币冲进央行金库，要求中国按面值兑换金条。
真正发生的，是石油、人民币和黄金三种资产之间的市场交换，而不是中国用国家储备黄金给人民币“兜底”。
近年来，中沙之间确实在为本币结算铺路。
2022年，中国提出利用上海石油天然气交易中心开展同海合会国家的油气贸易人民币结算；2023年，中沙两国央行又签署500亿元人民币、260亿沙特里亚尔的双边本币互换协议，为两国企业使用本币提供流动性支持。
这意味着人民币在中沙贸易中的使用条件越来越成熟，但能源定价和结算格局不会一夜翻转。
美元仍然是国际石油贸易的主要货币，人民币目前做的不是立刻取代美元，而是增加一条可以使用、可以投资、也可以退出的通道。
一笔人民币石油交易完成后，产油国拿到人民币，大致有几种去处：
购买中国机械、汽车、光伏设备和消费品；投资人民币债券、基金或产业项目；兑换其他货币；也可以进入黄金市场购买黄金。
黄金只是其中一个选择。
上海黄金交易所国际板以人民币计价，境外投资者可以使用离岸人民币参与交易。
交易所负责清算、交割和仓储，买方支付人民币，卖方交出黄金。买到的实物黄金还可以按照规则转运至其他国家和地区。
这里的关键在于，黄金不是凭空从中国国家金库里拿出来的。
市场中一方买走黄金，就必然有另一方出售黄金。出售者可能是银行、黄金企业、贸易商或投资机构，收到的则是人民币。
黄金的所有权发生了变化，人民币的持有人也发生了变化，但央行官方储备不会因此自动减少。
截至2026年6月末，中国官方黄金储备为7544万盎司，较5月末继续增加。
官方储备资产有独立的统计和管理体系，与境外机构在交易所购买多少黄金不是同一本账。
所以，“产油国用人民币买黄金会掏空中国储备”，本质上是把市场交易和国家储备混为一谈。
当然，这并不代表黄金大量外流永远没有影响。
假如境外需求短时间急剧上升，可能推高黄金溢价、改变市场库存，甚至增加黄金进口需求。
但这属于市场供求和价格问题，不是国家金库被人按固定价格兑换走了。
真正值得关注的，是产油国为什么愿意接受人民币。
过去，石油出口国收到美元后，可以购买美国国债、投资美元资产，也可以在全球市场自由使用。人民币要进入能源贸易，就必须给持有人提供同样清晰的使用出口。
能买中国商品，是第一层；能投资人民币资产，是第二层；能兑换其他货币或购买黄金，是第三层。
出口越多，产油国接受人民币的顾虑就越小。
从沙特的角度看，增加人民币结算不是要彻底抛弃美元，而是避免把全部能源收入压在一种货币和一套金融体系上。
它既需要美国的安全合作和美元市场，也需要中国的能源需求、制造业产品和产业投资。
多接受一种货币，就多一个谈判筹码。
中国得到的好处也很直接。
部分能源贸易使用人民币，可以减少企业反复兑换美元产生的成本和汇率风险，还能扩大人民币在大宗商品市场中的使用范围。
表面上是支付方式变化，本质上是在争夺国际贸易的金融入口。
但把这套安排吹成“人民币已经击败美元”，同样站不住脚。
国际货币基金组织数据显示，2026年第一季度，美元占全球已分配外汇储备的57.13%，人民币占比为1.99%。美元的市场深度、金融产品和全球流动性优势依然明显。
人民币国际化真正缺少的，不是一句“石油用人民币结算”，而是足够庞大、开放、多样的人民币资产市场。外国企业愿意收人民币只是第一步，愿意长期持有人民币，才是真正的突破。
能源问题同样不能夸大。
2025年，中国国内原油产量约2.16亿吨，原油净进口约5.8亿吨，其中从海合会国家进口原油1.8亿吨。
中国拥有庞大的采购规模、多元化进口渠道和较强的炼化能力，但对海外原油的依赖仍然很高。
这意味着中国可以增强议价能力，却不能单方面控制国际油价。石油价格仍受产油国政策、战争风险、航运通道、美元利率和全球需求共同影响。
真正可靠的能源安全，不是相信自己能够压住油价，而是扩大进口来源、提高国内产量、增加储备，并加快新能源替代。
因此，这套机制最有价值的地方，不是“用纸币换石油，再阻止别人换黄金”，而是让人民币成为一个有进有出、能够循环的国际交易工具。
我支持扩大油气贸易人民币结算，也支持用黄金市场、债券市场和商品贸易为人民币提供更多使用出口；但反对把它包装成黄金担保，更反对制造“中国金库将被搬空”或“美元霸权已经终结”的情绪。
沙特即使拿人民币购买黄金，也搬不走中国央行的黄金储备。
真正决定人民币地位的，从来不是金库里锁着多少金条，而是中国能否长期提供全球需要的商品、技术、市场和可信赖的金融资产。
黄金只能帮助人民币打开一扇门，强大的实体经济和成熟的金融市场，才有能力让这扇门一直开着。#美军暂停对伊空袭，国际油价开盘大幅下跌 $CL All eyes turn to July 29. The FOMC meets with rates at 3.75% and, for the first time in a while, a genuine hawkish tail: Warsh's Fed has an open dissenter in Logan calling for a hike, sticky inflation, and an oil situation that only recently began cooling. This isn't a "will they cut" meeting anymore.
The setup matters more than the base case. Markets still lean toward a hold, but the risk is asymmetric: a surprise hike, or a hawkish hold with hike guidance, would hit a market positioned for eventual easing. Crypto's firm today (ETH ripping +4% to $1,960), pricing calm into the meeting. I'd respect the tail: the distance between "hold and dovish" and "hold but hawkish" is where volatility lives. Watching the statement and the dot plot, not just the rate.
DYOR.
#FOMCRateWatch #OKXOrbit🚀 $HYPE Real-time price updates 🚀
Received! The current price has climbed to $60.3, rebounding over 6.6% from the previous day's low of $56.56, successfully reclaiming the psychological $60 psychological level and the EMA 50 moving average. The 24-hour high has been refreshed to $60.85, with bulls testing the last vacuum zone before the EMA200 ($62.49).
📊 Support and resistance levels (real-time correction)
Resistance Level (Resistance Above)
· $60.85 - $61.20: Immediate firepower point formed by the intraday high and the 4-hour upper Bollinger band
· $62.49 - $63.21: EMA200 and the key resistance zone highlighted by Gate analysts; a breakout would open upside
· $66.79 - $76.67: Fibonacci 0.618 with a previous all-time high forming a hard-top trap zone
Support Level (Lower Defensive Line)
· $59.50 - $60.00: The original resistance level has turned into first support, and the $60 level has become a bullish stronghold
· $57.80 - $58.37: Pivot point and 4-hour mid-band, a buffer zone during pullbacks
· $56.00 - $56.56: Recent low; a break below would signal a false breakout
🐋 Market maker movements on the chain
While the price broke through $60, whales that withdrew and staked $32.87 million on July 25 showed a floating gain of over 10%**. The entity's nine-month average price is $44, and the current book is highly profitable. Total staking volume climbed to 436 million, with lock-up reducing market circulation selling pressure.
⚠️ However, nearly 2 million Multicoin Capital tokens ($120 million) have been unstaked, and the pressure from a16z-linked addresses remains at a peak. On July 27, 20,640 tokens (about $1.24 million) were burned in a single day, with deflation continuing to absorb selling pressure.
📈 Positive factors
Successfully breaking through the $60 psychological barrier triggered quantitative buying, putting pressure on futures bears. If it holds above $60, the technical side will confirm a bullish bottoming structure.
Hyperliquid's net income reached $800 million, with cumulative buybacks and burns of 4.73% of circulating supply. HIP-4 Upgrade and Prediction Market Launch continues to empower the ecosystem.
The $600 million unlock on July 29 has already been partially priced in by the market ahead of time, down 22%. If the unlock stabilizes above $60 with increased volume, it would signal a bearish exhaustion reversal.
📉 Bearish factors
July 29 coincided with the FOMC decision—$600 million unlocking flood combined with hawkish risk, high-beta counterfeit faces a double macro + supply shock.
Large unstaking and transfer records between Multicoin and a16z show that primary market capital is still exiting, raising doubts about market makers' willingness to hold onto the position.
Currently, the RSI is approaching 58, and although the CMF has improved, it has not turned positive. The price broke through $60 but trading volume did not significantly increase, indicating a risk of chasing the price above the volume and price warning.
⚠️ The above analysis is based on publicly available on-chain and market data and does not constitute investment advice. The $60 battle will be decided tonight. The July 29 market change window is approaching, so strictly control position risk. 🎯 $HYPE #长鑫科技上市, global storage competition adds variables #美联储周四凌晨公布利率决议 #美军暂停对伊空袭, with international oil prices sharply falling at the open [Graphic Observation | US Dollar Liquidity] At 19:16 Beijing time, Jin Ten article clues: Bank of America warns: The weakest window for US stocks is approaching, and gold and the dollar are expected to present opportunities.
Forex snapshot: EUR/USD 1.1392 (+0.20%); USD/JPY 163.63 (-0.14%); USD/RMB 6.7662 (-0.06%). The dollar's direction affects global liquidity pricing and also alters the sensitivity of risk assets like BTC and ETH to news updates.
Background summary: Does the 'August curse' really exist? Market data from the past decades is reminding investors that the summer rally is not simply a "falling season." Bank of America analysts point out that from August to October, a defensive market environment is more likely to emerge, with the dollar, gold, and bonds likely to be more favored by capital than stocks.
Verification point: If USD/JPY continues to rise and offshore RMB is under pressure, risk appetite may become more cautious; If the dollar falls, the resilience of risk asset rebounds is even more worth observing.
Risk warning: Central bank interventions, tariff news, or sudden geopolitical events may alter the exchange rate's transmission path. For market observation purposes only and does not constitute investment advice.I added another 64,000 to buy 10,000 USD, so in 2 days you can get 47 USD.
When you feel Bitcoin's price has become cost-effective, doing dual coins is also a good idea.$AEON Focusing on this project: a project incubated by Binance YZi Labs. Binance Ventures and ZI Labs are already high-risk, and the entire team is Chinese—some have worked at Binance, some are ordinary engineers at Google, and some have never led a team to develop Google's core projects. Just Li Yiyang, an ordinary engineer on the sidelines of Google, is hyped up for a long time. What is this project about AI payments, how does it connect to global merchants? Nothing has been realized. It's just a pipe dream project. Speaking of projects connecting to real-world merchant payment scenarios, the wallet was already established back in 2021. Other wallets had crypto bank cards and had already connected to dozens of merchants, integrating online payments and offline collections, but within two months, the project failed, proving that this kind of merchant-connected payment scenario track simply couldn't take offCrude oil fell nearly 9% in a single day, but the crypto market actually strengthened
A few days ago, the market was still worried about rising oil prices, but today the direction suddenly reversed
Brent crude fell below $85, with intraday losses widening to 8.77%
If low oil prices can be maintained, transportation and production costs may decrease, easing market concerns about inflation
BTC rose 0.78% today, ETH gained 2.6%, and risk sentiment has temporarily recovered
However, the sharp drop in oil prices may also be due to weakened demand expectations, so it should not be interpreted solely as positive
Going forward, it will depend on whether crude oil can stabilize and whether US Treasury yields and the US dollar will both fall $BTC $ETH #英伟达拟为OpenAI提供2500亿美元担保
In the past couple of days, many friends working in US stocks and Web3 AI have been discussing a phenomenon: NVIDIA is extremely actively investing in its own customers (such as CoreWeave, Lambda Labs, Mistral AI, etc.). After these startups secure financing, they immediately spend billions of dollars in orders to buy NVIDIA GPUs; Even more extreme, NVIDIA has provided credit guarantees or order binding for these customers' hashrate debts in some form.
Some say this is the most perfect closed loop of capital ecosystem in history; others say it is an extremely dangerous false prosperity. You ask me what I think?
To be honest, as a trader who has experienced several bull and bear cycles in the market, I first saw this pattern and felt chills down my spine—because it was almost identical to Cisco's ruined "Vendor Financing" during the 1999 telecom bubble.
Let me start with my clear view: in the short term, this is an unbeatable moat for NVIDIA to squeeze out competitors and secure computing power dominance; But in the medium to long term, if the real revenue of underlying AI applications can't keep up with the capital expenditure of chip frenzy, this will be a nuclear bomb-level structural risk mine.
Why do I say this? Let's break down the internal logic of this closed-loop and take a look:
First, this "capital flywheel" heavily relies on end-user applications to generate real fiat cash flow.
NVIDIA invests in customers with one hand and recovers GPU payments with the other, seemingly seeing a surge in revenue and gross margin soaring to 70%+. But the chips these cloud vendors buy ultimately have to be paid for by AI startups and enterprise clients renting computing power. If OpenAI and Anthropic burn tens of billions of dollars but their models fail to monetize expectations, and computing power leasing demand plummets, this closed loop will instantly shift from a "positive feedback flywheel" to a "negative feedback spiral."
Second, history always rhymes with similar rhymes.
On the eve of the 2000 internet bubble, Cisco also created a revenue legend by providing huge financing guarantees to telecom operators, prompting them to frantically buy their own routers. But when terminal internet traffic monetization failed and operators went bankrupt, Cisco not only failed to recover billions in bad debts but also suffered years of inventory destocking nightmares, with its stock price evaporating nearly 90% overnight. Nvidia's cash flow now is indeed much more abundant than Cisco once was, but the risk contagion chain of "lending money to customers to buy their right-hand products" remains unchanged.
Third, this risk is already extending into the Web3 computing power sector.
Many DePIN and AI computing power tokens in the market are essentially the lowest sub-derivative of this supply chain financing—buying a few NVIDIA computing power servers and issuing tokens under the guise of decentralized computing power rental. Once the supply-demand balance between top-tier NVIDIA and Neo-Cloud cloud arises, these underlying computing power concept projects without real paying users will be the first to be squeezed out of liquidity.
How long do you think Nvidia's closed loop can last? At what point will the Capex frenzy in US tech stocks hit its monetization ceiling? Feel free to share your thoughts in the comments section.AI is a once-in-a-century technological revolution, but the current market is half real growth, half bubble blowup. Many people only see AI concept stocks surging wildly, but don't notice the risks accumulating behind them. Blindly chasing at higher prices is very likely to become the buyers.
So here's the question: is there a simple and intuitive way to help us determine where the turning point for the current AI market peak is? Today, we will thoroughly explain the truth about AI investment using three core logics.
---
First, AI's long-term certainty is real.
From an industry fundamental perspective, this AI wave is not just about hyping concepts, but about a supercycle created by real money. The Philadelphia Semiconductor Index, which represents industry prosperity, has doubled its cumulative gains in half a year, rising more than 20-fold over the past decade, making it the fastest-growing sector among all industries. Demand is tougher—the five tech giants—Amazon, Google, Microsoft, Meta, and Oracle—are expected to spend a combined $800 billion in capital expenditure by 2026, nearly doubling year-on-year.
From AlphaGo defeating Lee Sedol ten years ago to AlphaFold winning the Nobel Prize in Chemistry, AI has long moved from the lab into industry. The implementation speed of large models, autonomous driving, and industrial AI far exceeds expectations, with computing power demand growing threefold every year. In other words, there is no problem with the long-term logic of AI changing the world; this is the most certain industry trend for the next decade.
---
But we must also recognize the risks.
First risk: Right now, only shovel sellers are making money; those digging for gold haven't made a profit yet.
Behind the boom lies a harsh reality—in the entire AI industry chain, the real profits currently come from upstream hardware vendors selling computing power. Ninety percent of global AI industry chain profits are concentrated in the hands of a few companies: Nvidia, TSMC, Samsung, and SK Hynix. Downstream cloud providers, large model companies, and AI application companies are basically throwing money to compete for the track, and their profit models have yet to fully develop.
For example: Google's AI-related capital expenditure this year reached $187 billion. To raise funds, it issued 100-year bonds with a coupon rate as high as 6%, but the revenue increase from AI is far from covering investment costs. Most importantly, AI has yet to produce a nationwide consumer killer application; most revenue still comes from corporate procurement and government projects, without forming a large-scale consumer market.
Upstream is already making a fortune, while downstream is burning money to tell stories—this structure is inherently unhealthy. This is precisely the most important industry logic for judging the turning point.
The second risk: rapid iteration, rapid capacity expansion, and the phenomenon of bubble differentiation imminent.
The AI industry has two unavoidable characteristics: most companies are destined to be eliminated.
First, industry iteration is happening too quickly. AI hardware updates only take two to three years. The computing power center built at 10 billion yuan today may become outdated in two or three years due to technological upgrades, resulting in massive investments being wasted and huge depreciation that could directly cut profits.
Second, capacity expansion is too rapid. Currently, global chip storage manufacturers are frantically expanding production, and SK Hynix has already started slowing the pace of high-end storage expansion, indicating that the industry has anticipated possible future oversupply. Once downstream demand growth can't keep pace with expansion, the situation will immediately shift from being scarce to a price war, with profits shrinking rapidly.
Take the 2000 internet bubble as an example—back then, anything connected to the internet could rise, but when the bubble burst, over 90% of internet companies delisted, and less than one-tenth survived. The current AI track is exactly the same: once the tide recedes, most concept stocks will show their true colors. Only leaders with real technical barriers and real performance can survive.
---
Earlier, I mentioned two risks—downstream not making money, oversupply on the supply side—both pointing to the same question: Can leading companies continue to hold up the valuation of the entire sector? If the faucet can't hold up, the entire chain will collapse.
Can the leading stock really hold up? This is where inflection point judgment comes into play.
Bank of America strategist Michael Hartnett provides a very intuitive AI benchmark for judging market turning points, so you don't have to guess blindly:
The MAGS ETF, which tracks the seven AI giants in the US market, is the core indicator of the entire AI market.
· If MAGS falls below $65, it means the leading stocks can't hold on, and the entire supply chain of storage, semiconductors, equipment, and materials will likely come under full pressure. At this point, be alert to risk spread—reduce positions where you should, and run away where you should.
· If MAGS holds above $70, it indicates market sentiment is recovering and capital is flowing back, signaling a relatively stable re-entry.
Why can an ETF act as the master switch? Because the core logic behind making money in the entire AI industry now lies in the seven giants, who are the engine of the entire industry. When the engine fails, limbs suffer. So don't just focus on individual stocks; focus on MAGS—65 and 70 are the lifelines of the industry.
---
To sum up: the AI supercycle is fine, but in the short term, it has already entered the bubble differentiation stage. In the long run, areas like computing infrastructure and domestic substitution must be supported by real performance; small stocks that only tell stories without core technologies could burst at any time.
Don't blindly chase highs in investment; look for companies with real cash flow and technological moats, and don't buy into vague expectations. Remember the MAGS anchor point—at least you won't be in the dark when the turning point comes.
The above are personal views and do not represent investment advice. Please be aware of the risks.#以太坊验证者退出队列已降至零
The number of people waiting to exit Ethereum staking has been reset to zero. Simply put: if you want to withdraw your staked ETH now, you don't have to queue, can leave anytime, and doesn't have to wait long.
Looking back to last year's market downturn, a bunch of validators swarmed to redeem and exit queues, blocking over 2.6 million ETH. Withdrawals took more than forty days. At that time, everyone panicked, afraid big players would crash the market, and the selling pressure looked terrifying. Now, no one is lining up to exit; essentially, large and retail pledgers don't want to cash out and leave, and confidence in long-term trends is returning.
Here's an interesting contrast: they exit instantly, but if you want to enter, stake and deposit, you have to wait 43 days. Over two million ETH are blocked at the entrance, waiting to get in. No one wants to run, while others rush in to claim staking yields—supply and demand are obvious at a glance.
From the perspective of ordinary crypto traders, this data is a solid positive sign, meaning there is no short-term risk of concentrated unlocking and sell-offs, and selling pressure has basically bottomed out. But don't get carried away blindly; short-term prices depend on the Fed's rate cut pace and the vote on the US crypto bill CLARITY.
In the long run, a large amount of ETH is locked up as collateral for a long time, so fewer coins circulate in the market, scarcity is gradually rising, so those holding long-term can rest assured; For short-term traders, don't rely solely on this data all-in; just put macro news first.Today, the China, US, and South Korea markets are likely all focused on Changxin's IPO. Although I don't trade the large A-shares, since it concerns my rebound positions in Hynix $SKHY and Micron, I must pay close attention.
The importance of Changxin's IPO has been extensively reported by various self-media, so everyone should be familiar with it:
1. For the large A-shares, there is now a flagship stock that can benchmark against the hottest memory sectors in the US and South Korea.
2. Regarding the China-US AI competition, the financing model has upgraded from government-led support to a joint financing involving government capital, industrial capital, bank credit, and public capital, opening the ceiling for commercial capital circulation.
3. The previously feared capital drain in large A-shares did not occur; today, the A-share indices closed broadly higher.
4. Although Changxin still has a technological gap compared to Hynix and others, China's recent years of overtaking and surpassing in multiple fields such as automotive, high-speed rail, power grids, photovoltaics, and rare earths have made industrial sectors in various countries shudder. Although the market generally believes there is still a three-year gap in HBM technology between China and South Korea, the pressure from the advancing "steamroller" chasing after them and the sense that catching up would kill the competition has impacted the stock prices of Korean and US giants. The path of inflating valuations by storytelling has been further blocked.
5. US capital is not monolithic either; Apple has repeatedly lobbied Trump to approve the use of Chinese memory in products sold in China. If realized, this would be a huge credit endorsement for Changxin's market acceptance. It would also significantly increase the profit margins of Apple's already price-increased products, which is one of the reasons for Apple's recent stock surge.
6. Changxin's IPO is somewhat analogous to SpaceX's, featuring a small float (6.73%) plus strategic high premium. Because the issue price was set relatively low, media outlets are now overwhelmingly promoting the first-day rise of 466% and a market cap exceeding 3 trillion. However, for those of us currently experiencing SPCX's halving, it is clear this implies potential shorting opportunities later. Yet, shorting large A-shares is technically difficult, so finding opportunities to go long on Hynix and Micron later is also a form of hedging. $SKHYNIX $SKHY $MU Changxin Technology's stock price surged nearly 5 times! How much higher can it still go?
Changxin Technology (688825) closed at 49 yuan on its first day of listing on the STAR Market, soaring nearly 5 times above its issue price!
Its total market value surged to 3.28 trillion yuan, surpassing Industrial and Commercial Bank of China in one fell swoop, topping the A-share market value rankings.
As China's first and the world's fourth largest DRAM manufacturer, the company’s global market share reached 7.67% in Q4 2025.
Comparison of the four major leaders:
Samsung: 33.96% share, market value about $1.12 trillion
SK Hynix ( $SKHY ): 34.48% share, market value about $1.13 trillion
Micron ( $MU ): 23.41% share, market value about $1.04 trillion
Changxin Technology: 7.67% share, market value about $484.7 billion
From a market value perspective, a reasonable imagination range for benchmarking against global leaders is roughly 1.5 to 2 times, which could approach or even briefly surpass the three storage giants. But its share is almost impossible to catch up with Samsung and Hynix.
The company's net profit attributable to the parent company is expected to be 50-57 billion yuan in the first half of this year, with production capacity expanding from 270,000 wafers/month in 2025 to 450,000 wafers/month in 2027.
Combined with increased procurement from domestic cloud and mobile phone manufacturers plus the explosive demand for AI server DRAM, the long-term market share is expected to reach 17%.
However, the strong cyclical and highly volatile nature of DRAM remains unchanged. Current prices are at a high level, and whether supply and demand support can continue remains uncertain.
Industry status, growth elasticity, and cyclical risks coexist.
Do you think Changxin Technology's current valuation is reasonable? $BEAT Perpetual contracts maintain negative rates and increase in open interest, with the core contradiction being the mismatch between strong derivatives short-short momentum and insufficient spot FDV support.
The price broke through $4.19 to hit a 24-hour high, with $4.167 billion FDV and $309 million in circulating supply forming a clear liquidity structure tear. Derivatives market open interest rose to $13.49 million, accompanied by a sustained negative rate of -0.08% to -0.05%, confirming that short stop-loss liquidation was the direct liquidity source driving the 24-hour 22% increase.
The liquidation chain in the derivatives market ranks first among the drivers, with negative rates forcing short positions to convert into buys; Spot funds attracted by repo and burn mechanisms and AI concepts rank second, providing initial liquidity for a breakout.
The upside scenario requires perpetual open interest to continue breaking through $13.49 million, while the funding rate remains deeply negative below -0.05%, effectively pushing the price to hold above $4.20. If the funding rate rapidly narrows toward the zero axis, passive short squeeze buying will lose support.
The downside scenario is triggered when the $4.20 high faces heavy spot selling pressure and the rate returns to zero. Since only about 30.9% of the total supply of 1 billion is in circulation, once derivatives are cut off, the lack of spot support can easily trigger a deep pullback toward the 24-hour opening price of $3.44.
When open interest drops sharply and funding rates turn positive, the benchmark assumption for short squeeze drivers fails, and the market returns to the spot supply-demand game phase.
The most important variable to watch in the next 24 hours is whether the $13.49 million opening interest can continue to expand, and whether the funding rate will tighten sharply.
#美国禁止开源AI的预期大幅回落 #AFX跨链桥被盗2415万USDC #多数党领袖称CLARITY休会前难通过CORE's 'grandma going down the stairs' wave, 0.024 is still pushing down—who's taking the lead? Looking at CORE's daily chart, it really looks like the elderly lady downstairs at my house walking down the stairs—pausing with every step, occasionally missing a step. Latest market data: CORE current price fluctuates around $0.024, down 1.4% in 24 hours, down 5.6% in two weeks, down 9.7% in 30 days. The drop may seem small, but looking back, it's a disaster—it's already pulled back 99%+ from its 2023 ATH of $6.14, with its market cap shrinking from a peak of nearly $10 billion to around $30 million, and the market cap ranking has dropped below the top 500. That real 'short streak' at the end of March was the real 'short sell-off': a single-day drop from $0.13 to $0.03, a 50% drop in one day, and a trading volume-to-market cap ratio soaring to 380%—this wasn't a swap, it was a panic escape. The trigger was the chain liquidation of the colend ecosystem lending protocol, combined with the forced consensus upgrade at the time that decoupled a large amount of CORE and delegated computing power. Validators were adjusting their own resources, with no one taking over. Why can't the old lady come downstairs for so long? Six kan stacked together: 1. BTC crash—CORE follows the BTCFi narrative, with a correlation coefficient of 0.87 with BTC. Bitcoin is grinding between 64,000 and 69,000, so knockoffs have no way out; 2. Unlocking selling pressure—total supply 2.1 billion, released linearly to 2137, only 59% circulating, 41% waiting to be released; 3. Chip concentration—90% of supply is held in 10is $MET the beta play to JUP?
does it outperform when $JUP rallies, and bleed harder when it doesn't?
I think at the moment they are correlated due to the nature of the assets and projects, so wondering if it makes sense to hold more of one or the other
would you rather hold $MET or $JUP ? BitMart exchange has also started running. In 2014, Mentougou closed down. Bitcoin crashed directly. 2017. Gathering coins and setting them out to sea. Binance is on the rise. Coin trading collapsed. The owners of the subsequent Ju coins have changed one after another. It's all about scamming. Previously, FTT went bankrupt. Only later did Merkle provide the Tree Asset Certificate. AAX, Zhongbian, and A.com all collapsed in one wave. Fcoin, once the world's largest trading and mining company, has also gone bankrupt. There is nothing that cannot go bankrupt. No exchange cannot go bankrupt. This wave has already listed many exchanges. Jubi flees to defend his rights. All exchanges have the possibility of collapse. The 10/11 incident. It was a bug at Binance. All market makers stopped placing orders. None of them engage in market making. A vacuum has appeared in the market. The market can drop all at once. You can also soar all at once. At that time, Binance bought orders and a series of staking bugs occurred. This led to a direct market crash. Market makers without risk control can also be directly taken advantage of. Binance was also hacked in 2020. No existing exchange is absolutely safe. Diversifying chips is indeed about spreading riskCZ: "Tough times again. Self custody..." The market heard "self custody". I heard "again". Bear markets aren't a single event. They're erosion. 2026 YTD: - Crypto market -25% - $BTC tested $64k - 99+ projects delisted or dead - 2 exchanges shut down in one week No headlines. Just quiet closures. The cycle repeats annually: Exchanges fail → tokens go to zero → new capital buys dips → repeat. My P&L leak wasn't bad calls. It was overtrading chop. Profit turned into bleed by being in the market evBEAT suddenly surged, and many people's first reaction was: Did the project receive some super positive news? But when looking at OKX's funding rates, long and short accounts, open interest, and trading volume together, it becomes clear that the core of this rally is not just news-driven. It is more like: buyback and burn combined with AI narrative laying the foundation, unlocking expectations that attract a large number of shorts to enter, price breaking through triggers continuous short covering, eventually evolving into position increases squeezing shorts. Let's first look at the latest data: BEAT latest price: about $4.19 24H open: about $3.44 24H high: about $4.20 24H low: about $3.39 24H change: about +22% 7-day change: about +76% 30-day change: about +104% OKX perpetual open interest: about $13.49 million CoinGecko circulating market cap: about $1.289 billion FDV: about $4.167 billion circulating supply: about 309 million BEAT total supply: 1 billion BEAT This is not an ordinary rebound, but a clear trend acceleration. 1. Why the sudden surge? First, shorts are crowded for a long time. OKX's funding rate has been negative for multiple settlement periods in the past, reaching: -0.08%, -0.07%, -0.06%, -0.05% during some periods. A long-term negative funding rate indicates that there are many positions betting on BEAT's price to fall. But the price never dropped as shorts expected; instead, it kept breaking through continuously. This forms a typical$AEON There is Air Coin. After much effort, OKX finally launched a new coin, but the storytelling Air concept of AI payments, integration with international malls, all of these were unfeasible Air Coins. These payments were made by wallets before, but they were outdated and not novel. Early wallet projects had crypto bank cards, on-chain payments, off-chain collections, and many functions beyond these implementations, but none succeeded or failedThe storage market may be moving from a "duopoly era" to a "three-way competition"
Over the past year, one of the biggest beneficiaries of the AI computing power wave has been the storage industry.
But the market has long held a default perception:
High-end storage is a game for Samsung, SK Hynix, and Micron.
Now, this pattern may be changing.
Changxin Technology was sought after by capital on its first day listing on the STAR Market, with its market value once reaching the 3 trillion yuan level, becoming one of the highest market value companies in the A-share market.
Just before this, Anthropic signed long-term supply and strategic cooperation agreements with Samsung and SK Hynix, with AI large model companies' orders continuing to concentrate on leading storage manufacturers.
An interesting phenomenon has appeared:
The capital market is still debating whether AI investment is overheated, but the industry side is continuing to increase investment.
My judgment is:
Changxin's listing may be more about valuation reappraisal in the short term, but in the long term, it will indeed change the global storage competition supply pattern.
The reason is simple.
The storage industry is not only about technological leadership but also about capacity, cost, and customer binding.
Samsung and SK Hynix's biggest advantage currently is their first-mover advantage in HBM and high-end AI storage fields.
But the significance of Changxin is that it officially brings China's storage capacity into the global capital market, adding an important variable to future DRAM competition.
For investment, I would not chase the sharp rise on the first day of listing directly.
Because a correct industry trend does not mean short-term prices are reasonable.
AI chips and storage stocks have already experienced a round of valuation expansion; what really determines the market later is not the story but whether orders can be continuously fulfilled.
If AI data centers continue to expand in the next few years and storage demand keeps growing, the market may accommodate more players.
But if AI capital expenditure starts to slow down, intensified competition may instead squeeze profit margins.
So now I focus more on three indicators:
First, whether HBM order growth continues;
Second, whether the storage price cycle reverses;
Third, whether the profit margins of various manufacturers can keep up with the expansion speed.
I believe the storage sector is not over yet, but the logic has shifted from "who has the AI story" to "who can make money in the AI supply chain."
The emergence of Changxin does not necessarily mean Samsung and SK Hynix lose their advantages, but it reminds the market:
In the AI era, computing power competition ultimately is not just about GPUs but the entire supply chain behind them.
#长鑫科技上市，全球存储竞争添变量
$SKHYNIX $SAMSUNG $MU #美联储周四凌晨公布利率决议
At early Thursday Beijing time, the Federal Reserve's FOMC interest rate decision, dot plot, and chairman's press conference will be released. This is the biggest recent macro event, directly determining the overall tone for risk assets in the next 1 to 2 months. Many newcomers only focus on "whether to raise rates or not." To be blunt, the interest rate remaining unchanged is basically the market consensus. What truly affects the market are the wording, the dot plot, and statements on inflation and the pace of rate cuts.
1. Current Market Background
1) The benchmark interest rate is currently in the 3.50%-3.75% range, and the mainstream market expectation is to keep rates unchanged this time;
2) The Middle East conflict has pushed up crude oil prices, and rising energy prices will bring inflation rebound pressure, so the Fed is reluctant to easily release easing signals;
3) The previous dot plot has already shown divergence, with half of the officials not ruling out restarting rate hikes within the year;
4) Bitcoin is currently in a range-bound consolidation, unable to break through the upper resistance, and funds are generally waiting for this rate decision result.
Three Scenario Simulations (directly corresponding to BTC market performance)
Scenario 1: Baseline Expectation (keep rates unchanged, tone neutral to hawkish)
Key wording: Inflation risk still exists, no discussion of rate cuts for now, possibility of future rate hikes reserved.
Market reaction: Short-term slight fluctuations, range-bound pattern continues. BTC continues to tug between 64000 and 66800, difficult to break out into a one-sided trend.
Scenario 2: Hawkish Surprise (keep rates, but significantly raise inflation expectations and revise dot plot upward)
Key signal: Clearly implies there are still rate hike options within the year, delaying rate cut expectations.
Market: USD and US Treasury yields rise, risk assets collectively under pressure. Bitcoin likely tests support at 64000; if effectively broken, further downside toward around 62000.
Scenario 3: Dovish Positive (keep rates, acknowledge inflation easing, release rate cut expectations within the year)
Key signal: Downplay inflation risk, release easing expectations.
Market: Risk appetite warms up, BTC challenges 66800 resistance; only by holding above can it open a new round of rebound space.
Important reminder: The market has long been pre-gaming expectations, making it easy to buy the rumor and sell the fact. Even if the result is dovish, if the positive impact is less than imagined, there will still be a spike followed by a pullback. Do not blindly chase longs.