Orbit Post Sitemap
[Risk Warning]
Macro level:
Geopolitical variables are stirring again, the likelihood of economic sanctions being implemented is increasing, and both the US and Iran verbally do not seek any so-called reconciliation. Neither side seems willing to back down in the struggle for control over the strait.
Oil prices have returned to 89, and the 30-year US Treasury yield has hit a new high.
The three major US stock indices are high, but this is mainly driven by individual stocks pushing the market up. If the momentum of this rebound weakens or pauses, it is highly likely that both the market index and individual stocks will pull back together, which may also lead to a retracement in the crypto market.
Structural level:
In the past two weeks, the drop was only about 3000 points, which is considered sideways movement on a larger scale. To reach the liquidity near 67500 upwards, it needs to hold above 65000, but judging from the results, we have failed many times consecutively.
The reason it can't hold is that the existing funds do not have consistent expectations, and the uncertainty is too high. Moreover, the recent rise is not driven by spot demand; the contract market's "aggressiveness" is ineffective.
Response strategy:
Personally, I have closed all positions in all accounts and am waiting in cash for direction.
If holding positions, set breakeven stop losses for long positions and do not open new ones for now. For short positions, try light positions with tight stop losses, such as today's 64500 short position, with a stop loss if it breaks above 65000.
$BTC $ETH $SNDK
#30年期美债收益率创2007年以来新高 BTC is back to 64,000: The most dangerous misjudgment now is mistaking "unable to fall further" for "about to take off"
BTC latest around $64,207, up 1.4% intraday, with a low of $63,246 and a high of $64,507.
It seems to be recovering, but the structure remains unchanged:
There are buyers around 63K, but no one chases near 64.5K
This means the current pattern looks more like a "bottom support + top profit-taking" box rather than a new upward trend.
The macro environment also limits breakout potential. The US 10-year Treasury yield has risen to about 4.74%, the 30-year yield reached 5.326%, near a 20-year high; Brent crude oil has also climbed back above $91.
Although gold fell back to about $4,397 today, it has still risen about 10% in the past month, indicating that defensive capital demand has not disappeared.
So I’m not in a hurry to call "bottoming complete"
There are only two real confirmations:
Volume-supported hold above 64,500 → trend recovery;
A renewed break below 63,200 → continued testing of lower support
The market indeed is increasingly unable to fall further, but being unable to fall is only a necessary condition for bottoming, not a sufficient condition for rising.
The last missing step now is incremental capital. $BTC
#30年期美债收益率创2007年以来新高 Many people, upon seeing the word "regulatory," immediately react: It's going to drop! On the contrary! What truly deserves attention now is that the US and Hong Kong are gradually pushing cryptocurrencies out of the "gray area" into the formal financial system. On the US side, the SEC's stance has clearly changed recently. Previously, they would deal with project teams only after problems arose; now they are considering establishing clearer rules for crypto projects and even researching specialized systems for crypto asset investment contracts. (Reuters) What does this mean? To put it plainly: it used to be "You go first, then I'll tell you whether it's illegal or not." Now it's turning into: "I'll write the rules first, you play by them." This is actually a huge boon for the entire industry. But the ruthless ones are still in Hong Kong. Hong Kong's stablecoin regulation has entered the practical implementation phase. On August 12, the first batch of Hong Kong dollar stablecoins called HKDAP was launched, but for now it is mainly aimed at institutions and professional investors, with plans to gradually expand to retail users in the future. (Reuters) So now I increasingly feel that this real major trend might not be a tenfold increase for a single altcoin. Instead: stablecoins + compliant trading platforms + traditional financial institutions + blockchain payments. In the future, those who truly benefit from the dividends may not be the projects that constantly shout "hundredfold coins." Instead, it might be projects that can enter banking, payment, clearing, and asset management systems. That's why I keep reminding you: don't just focus on the candlestick. Policy direction is sometimes more important than candlestick charts. BecauseBTC 冲上 64155，山寨却集体趴下——这画面很多人第一反应是"牛市来了"，但真相可能刚好相反。 你有没有想过，当大饼独涨、山寨普跌的时候，市场到底在害怕什么，又在买什么？ 我昨晚盯盘到凌晨两点，看着那根阳线拉起来，心里第一反应不是兴奋，是警觉。BTC 触及 64630 的 24 小时高点，ETF 净流入在续，降息预期在升温，大资金把 BTC 当避风港——这些逻辑都没错。但真正让我停下来思考的，是另一边那些跌得无声无息的币种。 WLD 昨天反弹 5%，今天直接吐回去 10%。CORE 昨天冲了 11%，今天原封不动还回去。BEAT 更夸张，从 0.37 被砸到 0.27，距离历史高点已经跌了 92%，而且还在继续出货。这些币有个共同点：它们昨天的上涨，根本没有基本面在托底，纯粹是短线资金找个位置做一波反弹。今天大饼一吸筹，这些浮筹立刻被抽干。 这里我想说一个很多人没注意到的细节：跨市场联动的真正信号，不在 BTC 本身，而在 HYPE 和 DOGE 的表现。 HYPE 59.3，涨了 0.57%，是今天唯一守住阵地的山寨。DOGE 只跌了 0.36%，几乎可以忽略。这说明什么？资金If the crypto world keeps playing like this, there will be no hope by the end of the year!
The current undeniable reality is the cyclical bear market, with neither funds nor enthusiasm here.
The first phenomenon is that stock assets dominate the trading volume rankings on cryptocurrency exchanges. As a user, would you rather buy $10,000 worth of air or $10,000 worth of SanDisk? Look at how crazy SanDisk is, with big volatility and so many stories to tell. The funds are still in the crypto world, but no longer in the coins. It's like the child is still yours, but already calls someone else dad.
For example, $CORE, how many times have you seen it drop? The highest price was 6.9, now it's 0.02, trapping so many people. That means if you bought $10,000 at the peak, it's now worth only $29—a total bloodbath.
Another example is cs coin $LAB, which peaked at 20 and is now 0.08. Countless players have been liquidated, pouring in their hard-earned money, hoping to leverage small amounts for huge gains and get rich overnight, but in reality, they're just toys for the whales.
The second phenomenon is that crypto gameplay is still so primitive: find a hot buzzword, then post a meme. Then pump the market cap. The current market cap is close to 100 million RMB, which is 100 times the actual box office. This reflects that the crypto world currently has no new narratives or rhythms.
Isn't the recent bull run a perfect example?
Exchanges have discovered a new world: if traditional assets can be bridged through cryptocurrencies, this is a user base orders of magnitude larger than the crypto world. So everyone is scrambling for traditional finance users—who the hell still invests in crypto projects?
$BTC has shown some signs of life these past couple of days, consolidating with low volume after so long. But I don't think this is a trend reversal; it might just be a struggle. Going forward, it will probably oscillate between 62,000 and 65,000, just shaking everyone up.
We acknowledge that crypto is in a tough phase, but I don't think crypto is done. This is just cyclical pain, and everything will revert to the mean. At most, two years—it's not like we can't wait. #财报观察员：小米即将发布财报，你更看好哪条业务线？ I personally am more optimistic about the automotive line. The smartphone segment is currently pushing towards high-end, and the average selling price is rising, but the pressure from rising storage chip prices is quite significant, heavily impacting shipment volumes and squeezing profit margins. On the automotive side, although it is still in the investment phase and may incur some losses in the short term, the delivery volume has been steady above 30,000 units for several consecutive months, and the annual target of 550,000 units doesn't seem completely out of reach. If the volume increases further and costs come down, this segment has a strong chance to become a true second growth curve. Very optimistic about Xiaomi Auto's future. $XIAOMI #闪迪收涨逾8%，长期协议受关注
I originally thought that after the short squeeze rally, the market would take a breather today, but it still closed up over 8%, indicating that the capital has no intention of leaving. $SNDK's current movement is no longer driven by ordinary positive news; the market is repricing it.
First, let's look at what the long-term agreement actually means.
A $93.9 billion long-term supply agreement is a major event for any company. For SanDisk, this is not just an order; it significantly reduces revenue volatility for the next few years. The storage industry used to fear cycles the most—big price swings causing profits to rollercoaster. Now, with a long-term agreement, it's like putting a safety net under performance. The market is willing to assign a high valuation, and that's the logic behind it.
But I also remind myself: an agreement is an agreement, delivery is delivery. No matter how attractive the targets from 2028 to 2030 are, they must be fulfilled step by step. The market can price in advance, but I cannot confirm profits on behalf of management prematurely.
What does this mean for us?
First, $SNDK's short-term sentiment remains strong. Closing up 8% shows profit-taking hasn't been widespread, and new funds are still flowing in. There may be short-term momentum pushing it higher, but the higher it goes, the greater the volatility.
Second, the storage sector will continue to diverge. $SNDK being strong doesn't mean $MU and $SKHY will also have big gains. Capital tends to cluster around leaders; follower stocks rise slowly but fall quickly. If $SNDK pulls back later, follower stocks will likely falter first.
Third, the aftershocks of the short squeeze are still ongoing. Many shorts were forced out earlier, but new shorts might re-enter at high levels. The tug-of-war between bulls and bears at these highs makes prices prone to large swings.
My personal view:
I previously shorted $SNDK and got squeezed out, so I'm extra cautious now. The long-term agreement is a solid positive, I don't deny that. But after the positive news has been priced in and the stock has risen so much, short-term sentiment is quite saturated. At this level, I'd rather miss out than chase higher.
What makes $SNDK most attractive now is the long-term logic, but the most dangerous aspect is the short-term slope. After a 35% rise in five days and another 8% today, such a steep increase rarely sustains historically. It will either consolidate sideways or experience a sharp pullback.
My approach:
Don't chase the highs; focus on pullbacks.
At this point, watching is more comfortable than chasing. The real opportunities are never when the price is surging the most.
$SNDK $MU $SKHY #闪迪收涨逾8%，长期协议受关注 #财报观察员：小米即将发布财报，你更看好哪条业务线？ Personally, I am more optimistic about the automotive line. The high-end trend in smartphones is progressing, and the average selling price is rising, but the pressure from rising storage chip prices is quite significant, which greatly affects shipment volumes and squeezes profit margins tightly. On the automotive side, although it is still in the investment phase and may incur some losses in the short term, the delivery volume has remained steady above 30,000 units for several consecutive months, and the annual target of 550,000 units doesn't seem completely out of reach. If the volume increases further and costs come down, this segment has a strong chance to become a true second growth curve. I am very optimistic about Xiaomi Auto's future.BTC's current rebound shows a notable change:
A few days ago, the price was fluctuating around $63,000, but today it has climbed back above $64,000; meanwhile, after several consecutive days of outflows, the US spot BTC ETF saw a net inflow of about $137 million on August 17.
This indicates that short-term capital sentiment is recovering, but we cannot yet equate breaking through $64,000 with firmly holding above $64,000.
I am more focused on the next few 4-hour cycles:
If the price can hold near $64,000 on a pullback without quickly falling back, this level may shift from resistance to support. The next short-term upward target I will watch is around $65,500 — which is the key level that needs to be truly broken in this rebound.
Currently, the main risk comes from the US stock market. QQQ is only about 2.3% below its high from the past year, and risk assets overall are at elevated levels. If macro data or the Federal Reserve meeting minutes turn out to be more hawkish than the market expects, US Treasury yields could rise, putting pressure simultaneously on tech stocks and BTC.
If the price falls back below 63,500–64,000 and cannot quickly recover, we need to be cautious of a return to $62,000.#财报观察员：小米即将发布财报，你更看好哪条业务线？
Xiaomi is about to release its earnings report. Which business line do you think is the strongest? $XIAOMI
My top pick remains the automotive sector.
Phones and IoT are Xiaomi's cash cows, but the automotive segment is the engine that could multiply its valuation several times. According to the latest data, quarterly total revenue is close to 100 billion, with automotive and innovation businesses contributing 19.9 billion. The SU7 and YU7 models have cumulatively delivered over 650,000 units. Once production capacity expands, the scale advantage will become evident.
In this earnings report, I’m most focused on the automotive business’s gross margin.
Delivery volume is relatively predictable due to capacity constraints, but the gross margin reveals the real strength. In Q1, Xiaomi Auto maintained a gross margin of 20.1% despite pressure, which is outstanding in the industry. If the gross margin can stay stable, it indicates strong cost control and directly dispels concerns about car manufacturing dragging down profits.
Long-term investment logic:
▶️ Phones and IoT:
Responsible for steady cash flow to fund R&D.
▶️ Automotive and AI ecosystem:
Responsible for raising the ceiling, turning car buyers into long-term users of intelligent driving and ecosystem services.
▶️ Forecasting the trend:
As delivery volume surpasses the breakeven point for high fixed costs, the automotive business’s profitability inflection point will arrive soon. Once the automotive segment starts generating net profits, Xiaomi’s entire valuation logic will shift from consumer electronics to a high-growth tech giant.
Are you more inclined toward the stable phone cash flow, or optimistic about the growth potential of car manufacturing? Let’s discuss in the comments~
DYOR [Pharaoh's Market Watch]
Pharaoh straightforwardly says that gold breaking above 4430 is not surprising; what truly matters is the shift in the options market's sentiment. Smart money is betting it can still rise.
Spot gold has broken through $4430/oz, with New York futures nearing $4490. The options market signals are even more worth watching: some investors bought about 8,000 November expiry call options with a strike price of 460 at around $5.55 each. The skew in gold options has shifted from "buy puts for protection" to "buy calls to speculate on a rise," while gold funds have recorded the strongest inflow since January. These three events happening simultaneously are leading indicators of institutions' increased risk appetite.
Gold and Bitcoin are taking completely different paths. Over the past year, gold has risen 31%, while Bitcoin has dropped 46%. Veteran trader Peter Brandt publicly mentioned considering selling Bitcoin to buy gold. Behind this are two entirely different capital characteristics—gold is treated as a safe-haven asset under macro pressure, while Bitcoin is still priced as a leveraged tech stock.
Regarding institutional forecasts, UBS expects gold to reach $5000 in the first half of 2027, while ICBC Standard Bank is more aggressive, predicting it will hit $7150 this year. The World Gold Council sees the second half of the year as a critical juncture.
The direction for gold is already clear. Bitcoin is still fluctuating around 63,000, and the seesaw effect with gold is becoming increasingly evident. Good trades are made by waiting. $ETH $BTC $SNDK #黄金站上4430美元，期权资金转向看涨 Nvidia’s role in OpenAI’s Ohio PORTS-Pike project points to a broader strategic shift: securing AI demand may increasingly require financing the infrastructure around the chips, not simply selling the chips themselves.
The planned site carries roughly 8GW of IT capacity, a 20-year OpenAI lease, $1.5B from Nvidia, and reported first-phase credit support of up to $105B covering leases, power and residual value. The re-leasing option may reduce some renewal risk, but it does not remove the complexity of long-duration capacity commitments.
My read: this structure could deepen Nvidia’s ecosystem advantage while making capital allocation and counterparty exposure more important to its investment case. Not advice, just analysis.
#NvidiaBacksOpenAIOhio#Goldman Sachs says the likelihood of a Fed rate hike in September is very low
"Goldman Sachs directly slams the table: Three major cooling data points are pushing the Fed's September rate hike into a dead end"
On August 17, Goldman Sachs Chief Economist Jan Hatzius released the latest research report, directly sentencing the Fed's September rate hike to death.
Wall Street hawks' bets were all misled by superficial appearances. Although the US unemployment rate dropped to 4.1% in July, this was due to a large number of workers leaving the labor force, with monthly potential new employment plummeting to 5,000, not even maintaining a fraction of the 50,000 breakeven line; after the spring consumption spree supported by tax refunds ended, real spending growth in the second half of the year was directly halved to just over 1%.
July's core PCE seemingly rebounded to 0.2%, but more than half of the increase was a false reading caused by asset management fees rising with the stock market, which will inevitably face downward revisions by the end of September. Among the twelve voting committee members, only about four or five truly favor a rate hike, and faced with the sharply cooling economy, there simply aren't enough votes to increase tightening.
Focusing on the early September nonfarm payrolls and August CPI as two critical lifelines, as long as there is no abnormal surge, the overly priced rate pressure will accelerate its retreat, the yield curve steepening and liquidity breathing room are gradually opening.
Don't be scared by market noise, calculate the liquidity bottom line clearly, and patiently wait for market makers to replenish positions. $BTC Just saw a guy going long TSLA with 20x leverage. You can tell just by looking at this kind of trade that he's really bold.
Asset: xyz:TSLA.
Leverage: 20x, direction: long, entry price: 334.15, position size: 72,881, quantity: 218.108.
To be blunt, daring to go this hard with 20x leverage means even a slight pullback will turn your face green. The biggest risk with this kind of trade isn’t just picking the wrong direction, but stubbornly holding on when you’re wrong—the more you hold, the worse it gets.
Whether you’re stubbornly bottom-fishing or emotionally overexcited, high leverage means almost no room for error. Don’t turn trading into a gamble on your life; the market won’t cut you any slack just because your position is big.
Cut your losses when you should, keeping some ammo is better than stubbornly holding on. Don’t wait until you really can’t hold anymore and lose even the chance to admit you were wrong. The US dollar has plummeted to a recent low, while long-term bond yields have surged to their highest level since 2007. This unusual combination has left many market watchers confused.
Many people habitually assume that high interest rates mean a strong dollar, but this time the logic has changed. The rise in US Treasury yields is purely due to a massive issuance of bonds that the market simply cannot absorb—in other words, there is an oversupply of government bonds. Meanwhile, the dollar is weakening because the market anticipates an upcoming rate cut, prompting funds to withdraw early.
This trend is a heavy blow to the stock market. Persistently high long-term rates hit tech stocks and high-valuation sectors the hardest, as the cost of financing rises and companies lose much of their confidence to buy back shares. With such attractive risk-free yields, large investors simply turn to passive income strategies, naturally draining liquidity from the stock market.#财报观察员：小米即将发布财报，你更看好哪条业务线？
Xiaomi is about to release its earnings report, and I've been reviewing its fundamentals these past couple of days. To be honest, I hold a small core position, and I didn't buy in early. So this earnings report isn't just for show for me; real money is waiting for answers.
Let's first look at what the market cares about most this time.
Xiaomi's biggest variable right now isn't phones or home appliances, but automobiles. Since the SU7 delivery, the market has been waiting to see if the losses in the automotive business can narrow and if the gross margin can turn positive. If the automotive segment continues to bleed heavily, overall profits will be dragged down. But if losses narrow beyond expectations, the market might reprice Xiaomi.
The phone business, in my view, is the foundation. Global consumer electronics are still in a weak recovery; Xiaomi's phone shipments are decent, but average selling price and gross margin are limited. This segment won't bring big surprises nor disappointments; it's all about stability.
IoT and lifestyle products have always seemed undervalued to me. Xiaomi's ecosystem moat is actually quite deep, and it's expanding overseas. Profit contributions here are relatively stable, but the market pays little attention because it's not flashy enough.
Internet services are Xiaomi's most profitable segment. Advertising and gaming revenues are cash cows, but growth ceilings are evident. If internet services grow faster than expected, that's a genuine positive.
What does this mean for us?
First, changes in automotive business losses will directly determine short-term stock price direction. If Q2 automotive gross margin remains negative but narrows significantly, the stock price might hold; if it worsens, selling pressure will emerge.
Second, Xiaomi's valuation needs the automotive story to support it. Without automotive, Xiaomi is just a hardware company, and valuation won't be high. With automotive, the market is willing to give it a premium for its "ecosystem." So automotive progress in the earnings report is key to maintaining valuation.
Third, internet services are the profit anchor. As long as this segment is stable, Xiaomi's overall profits have a floor. If internet service growth drops sharply, that's the real problem.
My personal view: I've held Xiaomi for over half a year, did a swing trade in between, and made some profit. Before this earnings report, I neither reduced nor increased my position. My judgment is: phones and IoT won't perform badly, internet services won't collapse, and the real explosive variable is automotive. As long as automotive losses narrow, the market will buy in.
But I don't expect an immediate surge after the earnings. For a large-cap like Xiaomi, it's more likely to trend after earnings rather than spike like a small-cap. So my plan is: if there's a strong open and rally after earnings, I won't chase but wait for a pullback; if there's a slight high open but then a drop, I won't panic as long as automotive data isn't particularly bad, and I'll hold on.
My approach: keep the core position unchanged, watch two data points: automotive business gross margin and internet service growth. If these meet or exceed expectations, I'll add a small position on pullbacks; if automotive losses widen beyond expectations, I'll reduce some first and wait for a cheaper entry.
Earnings reports are not something I gamble on. Managing position size to handle uncertainty is more reliable than guessing direction.
#财报观察员：小米即将发布财报，你更看好哪条业务线？
$BTC $ETH $SNDK
#交易之声：你的经验值得被听到 SanDisk crashed! Those chasing the highs are still out on the rooftop enjoying the breeze, while the smart money has long fled!
When it rises, it's an epic bull market; when it falls, it's like dancing on a grave.
SanDisk $SNDK surged 9% last night but plunged 5% pre-market today, completing a tragedy of chasing highs within a day. Western Digital, Seagate, and Hynix ADRs all dropped over 4%, and the entire storage sector went silent.
Three reasons tore the market apart:
Too much gain: A 600% surge this year, jumping from 970 to 1826 in less than three weeks, an 88% increase. Short-term funds made a killing and left pre-market; whoever takes the last baton is unlucky.
Macro pressure: The 30-year US Treasury yield soared to 5.31%, a 19-year high. The higher the interest rate, the less valuable risk assets become, with the storage sector hit first.
Institutions in a mass sell-off: Some are adding positions, others are liquidating—David Tepper's fund sold 280,000 shares outright, and Renaissance cut 99% of its holdings. The direction is completely divided, indicating huge disagreement here; it's not the time to bottom-fish.
Remember: It's natural for prices to correct after rising too much, but don't rush to bottom-fish when prices fall; first, see who is running!
Do you still hold $SNDK? Type "1" in the comments to unite and support each other! #闪迪收涨逾8%，长期协议受关注 $SNDK #30-year US Treasury yield hits highest since 2007
The 30-year US Treasury yield has risen to about 5.3%, reaching a high not seen since 2007.
What really deserves attention is not just whether the Federal Reserve will cut rates, but that short-term rate hike expectations have cooled while long-term yields continue to rise.
The supply of long-term US bonds is increasing, but the long-term capital willing to absorb them has not increased correspondingly.
Fiscal deficits, increased issuance of long-term bonds, reduction in overseas buyers, and massive financing by AI companies are all intensifying competition for long-term capital.
Therefore, the market is truly repricing how much extra compensation is needed to hold long-term US Treasuries, meaning the term premium is rising.
This is more important than simply discussing a single rate cut or a single CPI data point.
For $BTC and $ETH, the impact is also direct: when the risk-free rate stays around 5%, the opportunity cost of bearing high volatility risk rises, naturally putting short-term pressure on crypto assets.
So what really deserves close attention going forward is not just "when the Fed will cut rates," but when long-term US Treasury yields will sustainably turn downward.
That is more likely to signal that the funding pressure on risk assets is beginning to ease. BlackRock simultaneously increases positions in $BTC and $ETH: Wall Street is building a "dual-core Crypto portfolio"
Today's ETF data clearly illustrates the point. On the BTC side, 10 ETFs saw a net inflow of 722 BTC, about $83.48 million, with BlackRock IBIT alone inflowing 976 BTC, approximately $113 million, holding 749,945 BTC valued at $86.67 billion. On the ETH side, even stronger, 9 ETH ETFs had a net inflow of 25,218 ETH, about $109 million, but BlackRock ETHA alone absorbed 76,851 ETH, $334 million, holding 3,567,302 ETH valued at $15.49 billion.
Note a detail: the inflow into ETHA alone exceeds the total net inflow of all 9 ETH ETFs combined, indicating that other products are experiencing significant outflows; the money is not evenly distributed but is concentrating towards BlackRock.
But the key is not the numbers, it's the allocation logic. BlackRock buys BTC for scarcity, digital gold, and macro hedging; it buys ETH for on-chain finance, staking, and the growth elasticity of RWA infrastructure. This is not a bet on who wins, but rather putting both into the same portfolio: BTC for store of value, ETH for growth.
Wall Street may have already started treating Crypto as a divisible asset class, rather than an either-or choice.$BTC today at 64,134, up 0.55% in 24 hours, standing above 64,000 for the second time this week.
But to be honest with you, I’m not moving at this position.
First, looking at the capital flow: last week, BTC + ETH spot ETFs had a combined net inflow of $1.1 billion, ending the continuous net outflow since the start of the year. BlackRock’s IBIT took 80% of that. But ETF trading volume dropped to the second lowest since October 2024 — money has come in, but turnover hasn’t picked up. This means allocation funds are accumulating, not trading funds pushing prices.
Next, on-chain data: wallets holding 1000+ BTC hit a new high for 2026 on August 8. Whales are hoarding, not selling.
I’m thinking about one question: with ETF inflows + whale accumulation + an 85% chance of rate cuts, why is BTC still stuck around 64,000?
Because there’s no catalyst. The SEC meeting was canceled last minute, the probability of the CLARITY Act passing on Polymarket dropped from 82% to 19%, and the Senate won’t reconvene until September 15. All three paths are blocked simultaneously, so BTC can only stay flat.
Looking across assets makes it clearer: the S&P broke 7,800 to a new high, gold is consolidating at a high level of 4,375, and BTC is still grinding in the 63,000-64,000 range. The same inflation cooling script is playing out: US stocks run first, gold follows, and BTC moves last. This is exactly the same script as August-September 2024.
#BTC沉睡供应创新高，稀缺性再受关注 AI trading giants suffer monthly losses, and the most ironic part is: they might have been right about the era but lost due to their positions
Situational Awareness and the drawdowns related to AI bets have exposed the side of Wall Street it least wants to admit. An AI infrastructure story can hold up in the long term; but if you buy with high leverage, the market only needs one sharp drop to compress a decade of narrative into a few days of margin calls
I think this is more of a warning than ordinary losses
Many people discuss AI only asking about direction, no one asks about endurance. You believe computing power, electricity, storage, and models will change the world, but that doesn't mean your account can withstand those few violent crashes along the way. The trading world is ruthless; it doesn't reward being "ultimately right," it first checks if you can survive today
AI is not the problem
Leveraging faith is the scariest problem
#AI押注受挫，华尔街交易巨头月亏150亿美元 The opportunities and risks of $000660.KS lie in HBM: the closer to $NVDA, the higher the valuation, and the harsher the demands
SK Hynix's most core label now is HBM. As long as AI servers continue to expand, as long as $NVDA's GPU demand remains strong, and as long as cloud providers keep building AI data centers, $000660.KS will repeatedly be discussed by the market. It is not an ordinary DRAM company but a very critical piece in the AI computing power system.
The uniqueness of HBM is that it is not just "memory capacity" but directly affects AI chip performance. Model training and inference require massive data to move rapidly between GPU and memory; if bandwidth can't keep up, computing power is wasted. Thus, HBM has evolved from a semiconductor niche product into an indispensable part of the AI hardware platform. Whoever can reliably supply HBM can share the most core profit pool in AI capital expenditure.
$000660.KS's strength is built on this position. The market is willing to give it a high valuation because it is close enough to $NVDA, close enough to AI server orders, and close enough to high-margin products. Compared to ordinary consumer memory, HBM customers are more concentrated, certification is harder, supply is tighter, and prices are more resilient. For investors, this is more attractive than traditional storage cycles.
But the closer to core customers, the more concentrated the risk. $000660.KS's biggest advantage is its binding to core AI demand, and its biggest problem is also this binding. Once $NVDA adjusts its roadmap, once cloud providers slow capital expenditure, once Samsung or Micron catch up in next-generation HBM, or once customers demand price cuts, the market's premium for it will be quickly re-evaluated. The leader's premium is not permanent; it must prove it is still leading with every generation of products.
This is also why $000660.KS's market performance cannot be described as "HBM shortage, so prices keep rising." The real questions are: How long can the HBM shortage last? How much market share can Hynix maintain? What is the yield of next-generation products? Will capital expenditure expansion lead to future oversupply? These questions are more important than daily price fluctuations.
Additionally, if SK Hynix pursues a U.S. listing or ADR trading, U.S. capital will price it more directly. Many U.S. investors who want to buy AI memory have only been able to buy $MU, $SNDK, or related ETFs; if $000660.KS becomes easier for U.S. capital to trade, it will gain more attention and compete with $MU and $SNDK for the same pool of AI storage funds. Increased liquidity is positive, but valuation comparisons will be harsher.
Therefore, $000660.KS is now a typical core sector asset: strong story, real demand, high elasticity, but also very high expectations. The best narrative is not "Korean stock rebound" but "high-pressure pricing of AI computing bottlenecks." This position is valuable, but the market will scrutinize it with the strictest standards.
HBM has put SK Hynix on the main stage of AI and also means it can no longer be evaluated by ordinary cyclical stock standards. Xiaomi Financial Report Preview: If You Can Only Bet on One Line, Who Is the Real Trump Card for Restructuring Valuation?
With Xiaomi's latest financial report about to be released, the most heated debate in the capital market now centers on three core areas: Can high-end smartphones continue to erode market share? Can the automotive business support a second growth curve? How much synergistic value has the AIoT and Renchejia full ecosystem delivered?
If we extend the timeline to the next three to five years and look into Xiaomi's current underlying business logic, the answer is actually very clear.
High-end phones are the granary and foundation of the entire empire. It continuously provides Xiaomi with abundant operating cash flow, maintaining the system entry point for hundreds of millions of active devices worldwide. But it is undeniable that the entire consumer electronics market has long entered a red ocean of stock competition. While premiumization can significantly improve overall gross margin, it is difficult to reshape the company's overall valuation by an order of magnitude on its own.
The real key to breaking the ceiling and completely reshaping Xiaomi's valuation multiples is the automobile.
Smart cars have never been just simple means of transportation; they are physical carriers of mobile supercomputing terminals and cross-platform ecosystems. It has raised the lifetime value of single-customer customers from a few thousand yuan worth of digital FMCG to hundreds of thousands of yuan worth of luxury goods. Once the automotive business is fully established, Xiaomi will no longer be just a consumer electronics hardware manufacturer, but will truly secure a ticket to the top tier of global high-tech manufacturing and smart mobility.
In this soon-to-be-released financial report, the most noteworthy data is neither total revenue nor simple delivery numbers, but the slope of improvement in automotive gross margins.
With ample backlog orders, ramping up production capacity is only a matter of time. What truly tests Lei Jun's ability to integrate deep water into cross-industry industrial manufacturing and supply chains is whether the gross margin per vehicle can quickly turn positive and move closer to a healthy range of 15% to 20%. As long as the industrial flywheel of decreasing marginal costs begins to accelerate, the certainty of the automotive business moving from a period of high-intensity investment to a self-sustaining period will be greatly established.
For long-term investors, mature businesses provide a margin of safety, while new businesses unlock market value potential. When the "Human-Car-Home Full Ecosystem" forms a truly system-level closed loop, hardware sales are only the starting point for customer acquisition; continuous monetization through ecosystem collaboration, software subscriptions, and intelligent driving services is the ultimate confidence supporting a trillion-yuan market value in the future.
If you can only bet long-term on one direction—between high-end smartphones, smart cars, and the AIoT ecosystem—which business line do you most believe will support Xiaomi's next breakthrough? Which core data in this financial report do you most want to see that exceeds expectations?
---
The above content represents personal views only and does not constitute any investment advice. DYOR，NFA。
#财报观察员: Xiaomi is about to release its financial report. Which business line do you think is more optimistic? The latest data from August shows that inflation levels in the United States remain high, upstream product prices remain sticky, and the labor market has suddenly cooled again. Under these circumstances, the Federal Reserve is not in a position to cut interest rates in the short term. Coupled with the fact that the outlook for U.S. economic growth has not significantly improved, the duration of high U.S. interest rates may continue to lengthen. On August 12, the U.S. Bureau of Labor Statistics released data showing that the Consumer Price Index (CPI) in July rose 3.4% year-on-year, with the core CPI up 2.5% year-on-year. The year-on-year increase in related data slightly declined, but month-on-month growth resumed. On August 13, the Producer Price Index (PPI) showed a 4.7% year-on-year increase, remaining flat month-on-month. In terms of employment, nonfarm payrolls in the U.S. unexpectedly decreased by 23,000 in July, significantly below the market expectation of an increase of over 80,000. Additionally, the number of new nonfarm jobs added in May and June was revised downward by a total of 103,000. Some analysts believe the data indicates that U.S. inflation is still far from the 2% policy target, and that U.S. employment data has been extremely unstable in recent years, often showing sudden weakness. In this context, raising interest rates would inevitably increase recession risks, while cutting rates would be unfavorable to achieving inflation targets. Maintaining the status quo is the Federal Reserve's "least bad choice." On April 10, people were shopping for daily necessities at a supermarket in New York, USA. Photo by Zhang Fengguo (Xinhua News Agency). Changes in U.S. long-term government bond yields illustrate the issue more clearly. On August 13, the U.S. Treasury issued $25 billion in 30-year bonds with a winning yield of 5.216%, the highest since 2001. The 10-year bond yield issued the previous day hovered around 4.683%, also the highest since 2007. Currently, U.S. short-term rates have fallen somewhat due to easing CPI data, but long-term rates remain firmly pinned by fiscal supply pressures and term premiums. With the Federal Reserve shrinking its balance sheet and no longer purchasing long-term bonds on a large scale, the Treasury's rigid financing needs combined with persistent inflation expectations have kept the 10-year U.S. Treasury yield stable between 4.6% and 4.8%, while the 30-year yield has formed a new pricing center around 5.2%. The International Monetary Fund (IMF) recently raised its global inflation forecast for 2026 to 4.7%, explicitly stating that the "disinflation trend has stalled." High U.S. interest rates will inevitably transmit globally through financing channels. Research shows that the supply shock of U.S. debt leading to rising long-term government bond yields often drives other countries' 10-year bond yields to rise in tandem. The collective rise in global long-term bond yields is currently not unique to the U.S. This situation squeezes private sector investment space due to fiscal bond issuance, closing the window for loose global financial conditions. This photo, taken on September 17, 2025, in Washington, D.C., shows the Federal Reserve Board sign. Photo by Hu Yousong (Xinhua News Agency). Emerging markets and developing economies face dual pressures. On one hand, most emerging market sovereign debt is denominated in U.S. dollars, and high dollar interest rates increase pressure on local currency depreciation and sharply raise refinancing costs. Dollar bonds issued by these countries and regions during the low-interest-rate environment are maturing, and now must be refinanced at higher rates, causing interest expenses to soar. On the other hand, cross-border capital from non-bank institutions is highly sensitive to market volatility; once risk appetite changes, funds quickly flow out, exacerbating a vicious cycle of local currency depreciation and widening sovereign credit spreads. The IMF believes that emerging economies and low-income countries highly dependent on energy imports and lacking participation in technology supply chains are especially vulnerable: following U.S. rate hikes suppresses domestic demand; not hiking leads to capital outflows and imported inflation. Prolonged high interest rates expose weaknesses in developed economies. U.S. commercial real estate loans are approaching a maturity peak, office vacancy rates remain high, and combined with sustained high long-term rates, regional banks' balance sheets face significant pressure. Italy, France, and Japan have public debt-to-GDP ratios at historic highs, and rising long-term rates pressure banks holding domestic government bonds with declining capital adequacy ratios. Furthermore, according to the U.S. Congressional Budget Office (CBO), net interest payments by the federal government will exceed $1 trillion in fiscal year 2026. After interest payments consume much fiscal space, the traditional monetary and fiscal dual easing response model is no longer sustainable. Central banks dare not cut rates lightly, and governments find it difficult to increase spending. Once external shocks occur, the policy toolbox has far less "ammunition" than during the international financial crisis. Capital allocation directions are quietly changing. In a prolonged high-interest-rate environment, projects with longer investment return cycles such as power grids, nuclear power, green hydrogen, and large infrastructure become less attractive, forcing delays in capital expenditures for energy transition and reindustrialization. Valuations of long-duration assets like AI data centers and biotechnology may be compressed, and companies not yet profitable find it increasingly difficult to raise funds by storytelling, accelerating resource concentration in cash-rich leading companies. The pace of total factor productivity improvement slows, reinforcing the "low growth plus high interest rates" pattern. Pension funds and insurance companies, faced with high-yield risk-free assets, tend to shorten investment durations, reducing marginal buyers of risk assets and weakening market intrinsic stability. For the foreseeable future, the likelihood of U.S. interest rates remaining high for an extended period is high, inevitably affecting and changing the global financial environment. Therefore, all global parties need to face the reality of a high-interest-rate environment, abandon illusions of a "return to easing," and prepare to manage debt, adjust structures, and prevent risks in a relatively high-interest-rate environment. This is the most pragmatic choice at present. $BTC $ETH $COMP surged with a record-breaking governance budget, but the reality of the protocol's locked value dropping by 90% from its historical peak still looms over the market.
The spot market saw trading volume surge 14 times, pushing the price rapidly to a phase high of $18.53.
The DAO has just approved a $52 million budget and replaced the management team, preparing for a full shift towards institutional credit and real-world assets.
Retail funds attracted by the early liquidity mining model have been continuously retreating, forcing the market to price this huge expenditure as a prerequisite for switching to a compliance channel.
If the compliance whitelist and institutional asset integration can be implemented as scheduled and contribute real interest spreads, the restoration of risk appetite will drive new buying to continue expanding positions; if the compliance process is delayed, the rebound momentum will be hindered.
If market concerns about the security of the underlying smart contracts increase, or institutional admission reviews are excessively prolonged, short-term speculative positions may quickly flow back to safety.
If institutional risk control fails to pass on-chain admission in time, the current expected premium will be substantially falsified.
The key variable to watch in the next seven days is the strength of spot buying support in the critical price range after the volume breakout.
#闪迪收涨逾8%，长期协议受关注 #英伟达支持OpenAI俄亥俄AI工厂 #美国财政部推进GENIUS稳定币规则Retail slump, weak employment, inflation easing—three strikes in a row, and Goldman Sachs has spoken: a rate hike in September is very unlikely.
Hatzius puts it bluntly: with data this soft, what reason does the September FOMC have to raise rates? CME FedWatch shows the probability of a September rate hike has dropped to around 30%. The market was previously too hawkish; Goldman Sachs believes the pricing was off.
As rate hike expectations cool down, gold immediately reacts—prices are approaching $4430. Bitcoin also catches a breather, climbing back to around $63,300 after plunging when the rate hike probability surged to 82% in early August. $BTC is a high-beta risk asset; when rate expectations ease, its elasticity is even greater than gold’s.
What’s more notable is the signal from the options market. Quant giant Susquehanna points out that options money is turning bullish: investors are actively buying upside exposure on gold, while the one-month implied volatility remains near recent lows, meaning the cost to bet on a rise is very cheap. Someone bought 8,000 November call options on the GLD gold ETF at a strike price of $460—all indicating gold prices are expected to climb further.
Macro and capital flows are resonating: Goldman Sachs douses rate hike expectations → the dollar comes under pressure → real rate decline expectations → gold benefits, and risk assets like Bitcoin also get a boost; smart money in options is simultaneously betting real money on continued gains. Both sides are talking about the same thing: this gold rally might not be over yet.
Of course, Goldman Sachs leaves a caveat—unless August data dramatically changes in early September. But until then, traders are choosing to build up long positions first #30-year US Treasury yield hits highest since 2007
Let's get to the point. The rise in long-term US Treasury yields is no small matter anymore; the entire asset pricing anchor is shifting.
The 30-year US Treasury yield has surged to its highest level since 2007, and this signal must be taken seriously. A risk-free yield steady above 5% means what? It means that assets like $BTC, which generate no yield, naturally lose some of their appeal—you hold Bitcoin for a year, and unless you bet someone will buy it at a higher price later, you get no cash flow. Looking back at this year's performance, Bitcoin's maximum drawdown was 46%, while gold, which also carries an inflation-hedging narrative, actually rose 33% against the trend. Physical assets are hard currency in turbulent times, and in this regard, Bitcoin still can't compete.
The data is straightforward as well: crypto ETFs saw a net outflow of $90.55 million in one week, capital is indeed moving.
$ETH's situation is even more awkward. Don't be fooled by the 4% staking annual yield, which sounds reasonable on the surface, but with the 5.3% US Treasury yield right there, the opportunity cost of holding ETH is clear. Many hedge funds are already playing arbitrage—borrowing ETH to stake for 4%, then switching to US Treasuries to lock in 5.3%, pocketing the spread risk-free. Capital is always smart; it goes where certainty is higher.
Let's also talk about the recently hot $SNDK, which jumped from 1528 to 1786 in five trading days, a nearly 17% increase. Long-term supply agreements, 80% gross margin, and booming AI storage demand make this story sound attractive. But don't get carried away; the 5.31% risk-free yield is still there. With a current market cap of 266.2 billion and a dynamic PE of 23, in a high-interest-rate environment, the cost-performance ratio just doesn't work. More importantly, storage is fundamentally a cyclical industry, with supply and demand shaped by the expansion pace of Samsung, Micron, and SK Hynix. Sandisk alone can't control the market.
In summary, as long as the 30-year US Treasury yield doesn't turn down, crypto and cyclical growth stocks will struggle to have a large-scale systemic rally. Next, keep a close eye on the Federal Reserve meeting minutes for any dovish signals, and continue monitoring geopolitical developments.
At this stage, rushing in to bet on a rebound is really not worth the risk-reward ratio. Amid the BTC·ETH rebound, the real signal is coming from AI infrastructure. Is this a recovery of risk appetite unique to cryptocurrencies, or part of a broader capital flow? BTC has recovered to $64,000, and ETH has surpassed $1,900 again. Selective buying pressure is also detected in some altcoins. However, there is one point worth noting before interpreting this trend solely based on internal cryptocurrency factors. At the same time, in the U.S. stock market, SanDisk (SNDK) surged 8–11% in just one day, rising about 35% for the week. This followed the investor day of the AI infrastructure storage company, where long-term growth prospects were presented. The figures presented by SanDisk's management are specific: mid-to-high teens revenue growth rate by FY2030, approximately 80% gross margin on a non-GAAP basis, about 50% adjusted free cash flow margin, and a new order backlog of $93.9 billion. The key point is that the AI investment cycle is expanding beyond GPU semiconductors to storage devices, NAND, and overall data center infrastructure. This is related to the cryptocurrency market and ETH ETF has turned positive again, but I don't think it's time to shout "institutions are back" yet.
From August 4 to 7, the US spot ETH ETF saw net inflows for four consecutive days, totaling about $256 million; as of August 17, it has recorded about $5 million in net inflows.
The direction is changing, but the momentum is not strong enough.
What I’m more focused on now is not how much flows in on a single day, but whether there can be continuous buying for 5–7 days straight, with daily amounts in the tens of millions of dollars.
If the capital inflow continues to rise, this round of ETH will look more like a genuine institutional replenishment.
Do you think ETH is currently lacking funds, or is it missing a catalyst to accelerate capital inflow? $ETH #BitMine增持至581.5万枚ETH，质押率约87% Thoughts on the $SNDK Short Hedge Strategy
Recently, many Crypto funds are gathering to short $SNDK. If you want to profit from SanDisk's bubble squeeze but fear that the overall storage sector will continue to surge causing a naked short squeeze, going long 0.7x $MU + 0.3x $SKHY as a paired hedge is a highly cost-effective choice.
$SNDK's weakness lies in its overly single business focus, heavily relying on data center enterprise SSDs and NAND, with high sensitivity to cycles and currently a huge valuation bubble.
Meanwhile, Micron and SK Hynix's production capacity is more focused on DRAM and HBM (the underlying hardware for AI large model inference), which are more bottlenecked essential demand areas. Also, Micron has long-term forward contracts providing support, making its fundamentals stronger and bubble smaller.
The core of this hedge is to profit from the "bubble spread" between the two. When the market falls, Micron and SK Hynix are more resilient than SanDisk; when the market continues to rise, the long side can also lock in risk. Ultimately, this achieves a low-volatility, drawdown-resistant long-term holding advantage. #闪迪收涨逾8%，长期协议受关注 Many people have set their expectations for the bear market too high. These people generally have not experienced the 2022 bear market; they only know how the 2023-2025 bull market will be. A novice dog running to 50m is still not satisfied. You have to know that during the 2022 bear market, there were only three tokens around 100m in the entire second half of the year, and not many around 50m. Yet this bear market's volume in July and August alone has already matched the entire second half of the 2022 bear market. What more do you want? What more do you want? I still remain full of hope for the future. There will still be many opportunities on-chain, provided we survive the darkness before the dawn #30年期美债收益率创2007年以来新高 #高盛称美联储9月加息可能性非常低 The various benchmarks nowadays are confusing. Actually, there's no need to listen to all the hype from different parties. I think there's only one indicator to judge whether a closed-source new model is strong: whether its revenue has surged dramatically.
For open-source new models, there's only one indicator: look at the growth in the number of various community variant models and acceleration tools, and the speed of their iteration;
This week, the two open-source models that have been trending are Qwen-3.8-27B and Minimax H3, with downloads and various optimization tools skyrocketing.
As for those that have been released but no one cares about, not even discussed, they are definitely underperforming, no doubt! DeepSeek V4-flash is already a thing of the past, and the Pro version is currently quite popular but somewhat confusing. The GLM-5.3 model weights haven't been officially developed yet; we'll know the popularity by this week or next week;The ninth cycle of reincarnation!
Some brothers on the OKX planet have already started to exit slowly.
There is a vague feeling that the market is off. The on-chain active data of $CORE, such as daily active users and transaction counts, doesn't look bad. But the fact that the thirty-day fees are only $254.80 has become a public joke within the circle. Everyone clearly sees that there are transactions on-chain, but no money is being made, and TVL is still flowing out.
This is useless; it won't last long before it dies out. The market quickly returns to a lifeless state. Most brothers have seen through the truth behind this data and are embracing $BTC, waiting for the big cycle. Some choose $BICO, following tracks with real business revenue. Most have jumped out of this endless loop, leaving only me still holding on.
Having experienced the previous eight cycles—surges, declines, sideways moves, and false positives—I have seen it all. The principle is clear in my heart: without real revenue support, all on-chain data is just a castle in the air.
But human obsession is strange. Even after suffering countless losses, I still can't help but fantasize that maybe the next upgrade will break the current deadlock.
The mockery of others and the cold data haven't pushed me to make up my mind to leave. The cycle continues... #交易之声：你的经验值得被听到 #财报观察员：小米即将发布财报，你更看好哪条业务线？ $SOL **SOL is leaning bullish, $75.6**
SOL is gearing up for a big move.
After consolidating for a whole week, the RSI is at a neutral 52, hovering flat, with the 50EMA pressing down at $75.48. The price is grinding between $74.50 and $76. But the 4-hour chart is forming a descending wedge — once this pattern breaks above the upper trendline, the first target is $77.10.
The strongest signal: ETF net inflows have continued for 7 consecutive weeks, with $10.26 million last week, the strongest since May. Meanwhile, BTC ETFs saw outflows of $390 million, and ETH ETFs outflows of $2.26 million — institutions are quietly rotating into SOL. Bitwise BSOL bought $8.8 million in a single day; although only Bitwise and Morgan Stanley are active, the direction is clear.
Bluntz on X called a weekly-level bullish divergence, saying this is now a "boring accumulation phase." Agave v4.2 upgrade launches this week, and the staking ETF may get approved within weeks — catalysts are lining up.
If $77.10 is taken out → $78.1 → $88.7.
If $74.50 breaks down → $73.64 → $71.9.
The market is sideways, SOL is building strength. Don't chase at this level; wait for the wedge breakout confirmation before making a move. #30年期美债收益率创2007年以来新高 On August 18 Beijing time, the yield on the U.S. 30-year Treasury surged to 5.31%, breaking the pre-2007 financial crisis historical high. The 10-year Treasury yield also rose to 4.72%, causing a sharp shock to the global asset pricing anchor and triggering a chain re-pricing of global stocks, bonds, and commodities. Unlike previous Fed rate hikes that pushed interest rates up, this round of long-end yield increases is driven more by fiscal supply, long-term inflation concerns, and overseas buyer reductions, a market phenomenon called the "bear steep" trend, where long-end rates rise significantly more than short-end rates. There are three core reasons driving the yield surge. First, the U.S. fiscal deficit is high, and Treasury supply is flooding the market. U.S. federal debt is approaching $40 trillion, and the Treasury continues to issue large amounts of long-term bonds, greatly increasing long bond supply. However, market absorption is insufficient, and the auction market requires higher interest to attract funds, pushing up term risk premiums. Second, long-term inflation expectations persist. July CPI remains high at 3.4%, combined with Middle East geopolitical conflicts pushing up oil prices, the market worries inflation will not quickly fall back to the 2% target. For 30-year ultra-long bonds investors, inflation erodes fixed interest income, so they demand higher yields as compensation. Third, traditional large overseas buyers are passively reducing holdings. As the largest overseas holder of U.S. Treasuries, Japan, affected by its own rate hikes and yen exchange rate interventions, has been continuously selling long-term Treasuries to obtain dollars, massively withdrawing from the long bond market over the past few months; combined with multiple central banks' out$005930.KS is not as sharp as SK Hynix, but Samsung's real resilience comes from a “low expectation recovery”
Samsung has recently regained market attention, but its story is different from SK Hynix. $000660.KS is a high-purity HBM asset with a very sharp logic; $005930.KS is a complex giant, with storage, smartphones, foundry, packaging, panels, and consumer electronics. This complexity makes it less sexy, but also gives it another opportunity: as long as expectations improve from disappointment, the rebound potential will open up.
In recent times, the market has criticized Samsung quite a bit. The HBM pace is considered behind SK Hynix, the foundry business faces pressure from TSMC, the smartphone business is affected by the consumer electronics cycle, and the overall structure is too heavy to be as cleanly presented as a single AI supplier. Investors prefer purity, but Samsung is not a pure-play asset. This causes it to be easily overlooked by capital during the sharpest phase of the AI rally.
But capital rotation sometimes works like this. The first phase buys the purest, the second phase buys those not yet recovered. As long as AI storage demand remains strong, the market will eventually ask: with Samsung’s huge capacity, technological accumulation, and customer relationships, will it really always lag behind? If it gradually improves in HBM certification, advanced packaging, and customer onboarding, even just moving from “obviously behind” to “starting to catch up,” the stock price could be revalued.
Samsung’s advantages are scale and supply chain completeness. It not only sells memory but also does foundry, packaging, and end devices. The AI era increasingly emphasizes system-level capabilities, making storage, logic chips, advanced packaging, and customer collaboration important. If Samsung can integrate these capabilities, its potential is not necessarily smaller than that of a single storage company. The only question is execution.
This is also the most suitable narrative for $005930.KS currently: not the “strongest AI stock,” but the “low-expectation comprehensive tech leader’s recovery.” It doesn’t need to surpass SK Hynix immediately, just prove it hasn’t been left behind by the AI memory cycle. The market sometimes demands less from low-expectation assets; as long as bad news stops increasing and some good news appears, valuation can recover.
Of course, Samsung can’t rely solely on “cheap” and “recovery” to tell its story. If HBM can’t secure key customers for a long time, if foundry continues to lose ground, and if smartphones and consumer electronics drag profits down, it will continue to be shunned by capital. The advantage of a diversified giant is multiple growth points, but the downside is any one business can hold it back.
So when writing about $005930.KS, don’t portray it as a substitute for SK Hynix. It’s more like a different trade: Hynix buys leadership, Samsung buys recovery; Hynix buys purity, Samsung buys comprehensive resilience; Hynix buys HBM bottlenecks, Samsung buys a late revaluation after AI cycle diffusion.
In the AI rally diffusion phase, the sharpest assets aren’t always the most profitable. Sometimes, laggards just need to prove they’re no longer falling behind to get the market to bet on them again. A few words about last night's US stock market.
On Monday, the three major indexes all closed lower: the Dow fell 0.51%, the S&P 500 dropped 0.52%, and the Nasdaq declined 0.32%. Despite the consecutive pullbacks, both the S&P and Nasdaq achieved three consecutive weekly gains last week.
The biggest drag was geopolitical tensions. The US-Iran agreement negotiations stalled upon expiration, causing oil prices to surge—WTI rose 2.55%, Brent increased 2.65%. The 30-year US Treasury yield was pushed to 5.321%, the highest since 2007, which weighed on high-valuation tech stocks.
The seven giants collectively faltered: Microsoft fell 3.04%, Meta dropped 3.54%, and Nvidia, Apple, Amazon, and Google all declined.
However, the AI hardware sector showed a completely different picture. The Philadelphia Semiconductor Index rose 1.64% against the trend, rebounding over 21% from the late July low. Storage stocks exploded across the board: Kioxia ADR surged over 13%, SanDisk rose over 8%, Western Digital climbed over 5%, and Micron increased over 4%.
The market has split into two camps: the seven giants are suppressed by interest rates, while AI hardware is driven by orders. The FOMC minutes on Wednesday will reveal how intense the internal debates were, making it the biggest variable this week.
My position is not heavy; I will wait for the minutes before making any moves.
This is my personal view and does not constitute any investment advice.
$BTC $ETH $SNDK
#财报观察员：小米即将发布财报，你更看好哪条业务线？
#30年期美债收益率创2007年以来新高
#闪迪收涨逾8%，长期协议受关注 $SNDK dropped from 993 to 1835, a short-term increase of 80%. Those who chased above 1800 are trapped, while the shorts are celebrating wildly. I'm not siding with either—I’m waiting for a position.
At 1,676-1,516, I will start scaling in long positions and hold them firmly in the live market.
I’m confident enough to base my judgment on this move—not going against the trend blindly, but because I’ve identified the structure clearly.
First, looking at the technicals: RSI correction from 89 is inevitable, but EMA21 is supporting at 1,677, and EMA55 at 1,535. These two levels are the lifelines of this rally; the trendline hasn’t broken on the pullback here. "Breakout - pullback - confirmation" is a textbook-level buying zone.
Now the fundamentals: The $93.9B long-term agreement is not just a PPT slide; 8 customers have locked in 50% output for FY2027 and 66% for FY2028. The investor day showed gross margin at 80% and FCF margin at 50%—numbers that are industry-disruptive for storage. SK Hynix itself has warned of storage shortages in 2027; the supply side is tighter than you think.
Capital flow is even more direct: Renaissance cut positions by 99%, Appaloosa fully exited, that’s true, but JPMorgan resumed coverage with a target of 2,250, Goldman Sachs at 2,200, and 81% of analysts have a Buy rating—that’s also true. Smart money isn’t exiting; it’s rotating—short-term traders are leaving, long-term holders are stepping in.
So I choose to buy below 1,676, not bottom fishing for a rebound, but waiting to strike precisely after sentiment has fully recovered.
#闪迪收涨逾8%，长期协议受关注 #黄金站上4430美元，期权资金转向看涨 Everyone, gold has climbed another step. On August 18, spot gold broke through 4420 USD, standing above 4430 USD, with silver rising in tandem. Since the beginning of this year, gold has gained over 50%. This round of increase is no longer driven by short-term sentiment; several forces are pushing simultaneously behind it.
Bank of America’s Hartnett pointed out a key background: U.S. debt is approaching 40 trillion USD, and interest expenses continue to rise. When the world’s largest economy needs to keep borrowing new debt to repay old debt, gold’s appeal as the ultimate credit anchor is rising again. This is a long-term trend that cannot be resolved in one or two quarters; this is the deep logic behind gold’s sustained rise.
The capital side is also cooperating. Demand for gold options has shifted from downside protection to bullish options, and gold funds have recorded the strongest inflow since January. Squeeze’s statement is very straightforward: the trading side has shifted from “buying protection out of fear of a drop” to “fear of missing out chasing the rise.” This sentiment shift often means the trend is self-reinforcing.
For BTC, gold’s continued strength is not a bad thing. Both share the logic of easing expectations, but gold’s rise is more about long-term capital allocation, while BTC’s rise is more a short-term reflection of liquidity expectations. Gold holding this position supports BTC’s macro narrative, but whether BTC can keep pace depends on ETF capital and on-chain data cooperation.
What do you all think gold can reach this round? $XAU $BTC $ETH A September Fed hold may be the base case, but the more important signal is why that view has strengthened. July retail sales fell 0.6% month over month, nonfarm payrolls dropped by 23,000, and both CPI and PPI eased. Together, those readings point to softer demand, employment and inflation rather than one isolated weak print.
With market-implied odds of a hold near 69%, conviction is meaningful but not settled. My read: incoming jobs and inflation data now matter less for confirming the consensus than for testing how quickly it could unravel.
#GoldmanSeesNoSeptHike#财报观察员：小米即将发布财报，你更看好哪条业务线？ ——$XXIAOMI
Xiaomi released earnings after market close, holding the short position for now to see how it goes
Xiaomi's earnings report came out after market today, just in time. This short position was placed at 3.36, with the price hovering around 3.357 all day, so the unrealized profit is basically negligible.
Looking through market expectations, Xiaomi's Q2 revenue is about 108.8 billion, down roughly 6% year-over-year. The pressure on the smartphone segment is indeed significant, with global shipments down 26%, mainly due to actively controlling inventory to protect profits. The average selling price actually hit a new high, but rising storage chip prices are squeezing gross margins, with smartphone gross margin estimated to be just over 8%.
The automotive side is doing okay, with Q2 deliveries exceeding 100,000 units, maintaining over 30,000 units for four consecutive months. Cumulative deliveries in the first half of the year surpassed 180,000, showing a relatively stable overall trend. Automotive gross margin is expected to reach around 20%, with economies of scale gradually emerging.
However, new cars won't launch until September, so the real highlight is in the second half of the year. The pace of storage price increases is slowing in Q3, and with new car deliveries plus easing cost pressures, a fundamental turning point should emerge in the second half.
The short position is not heavy, so I'll hold it for now and wait for the direction. Bitcoin Bottom and Top Supply Ratio: 42% Loss Supply and Realized Price Support Line
As Bitcoin oscillates repeatedly around 64,200, the proportion of positions in loss within the overall on-chain supply has risen to 42.48%, indicating intensified market turnover.
Realized Price & Profit/Loss Supply Ratio: The realized price represents the average buy-in cost of all market participants, while the loss supply ratio refers to the proportion of positions currently at a negative return.
Accumulation of Loss Positions: Compared to profitable supply, loss supply has increased to 42.48%, reaching levels similar to past major bottom consolidation zones.
Stable Exchange Supply Ratio: The exchange supply ratio remains low at 0.14, indicating limited additional large-scale selling pressure.
Historically, when the loss position ratio exceeds 40%, it often signals a solid bottom formation after prolonged selling pressure digestion. 🟢🔴 European and American Contract Price Change Review｜August 18, 4:30 PM
🟢 Top 10 Gainers
Symbol Price Change Key Highlights
$SKDD 10.38 +14.95% 2x Short Hynix ETF, leading the market. Amid storage sector pullback, short tools are favored, but leverage decay is a major drawback.
$SOXS 41.59 +11.80% 3x Short Semiconductor ETF, chip stocks remain under pressure, this product is highly volatile and only suitable for high-risk speculation.
$OPN 0.05774 +11.02% Opinion small-cap pulse, trading volume only $6.09M, poor liquidity, significant large order slippage, low participation value.
$VVV 13.86 +10.39% Venice Token new coin hype, $5.31M turnover, short-term capital guerrilla warfare, no fundamental anchor, high risk chasing highs.
$GPS 0.017018 +7.98% GoPlus Security, $278M turnover but less than 8% gain, massive stagnation evident, extremely fierce long-short battle.
$OPG 0.10398 +7.81% OpenGradient, light trading ($2.29M), small market cap follow-up rally, easily manipulated.
$ALLO 0.29346 +6.42% Allora with $37.43M turnover, relatively independent trend but lacks sector effect, sustainability questionable.
$LIGHT 0.1802 +5.50% Bitlight, only $1.45M turnover, liquidity dried up, typical volume-less surge, limited reference value.
$OFC 0.00862 +4.93% OneFootball fan token, $4.18M turnover, small cap self-entertainment, few follow-up players.
$COMP 17.45 +4.80% Compound, established DeFi, $7.14M turnover, defensive rebound after oversell, no trend reversal signal yet.
🔴 Top 10 Losers
Symbol Price Change Key Highlights
$KORU 20.37 -17.86% 3x Long Korea ETF, severely hit. Weak Korean market plus leverage decay sharply magnify losses.
$ZHIPU 133.16 -14.73% Zhipu AI concept continues heavy pullback, $11.58M turnover, AI sector in retreat, previous gains fully given back.
$UP 0.37 -13.93% Unitas stablecoin concept drops amid volatility, $2.2M turnover, small cap with wild swings, no clear logic.
$SKUU 23.46 -13.56% 2x Long Hynix ETF, stark contrast to top gainer SKDD (short), both sides suffer.
$GALA 0.001419 -13.16% Gala gaming metaverse old coin, $8.49M turnover, no new narrative stimulus, follows weak market downtrend.
$RAM 13.23 -12.50% 2x Long DRAM ETF, storage sector sentiment reversal, leveraged products fall more.
$IRYS 0.01344 -11.98% Irys storage concept, only $870K turnover, liquidity nearly dried up, small sell orders can cause deep dips.
$MVLL 28.03 -11.88% 2x Long MRVL (Marvell) ETF, dragged down by US chip stocks, leverage amplifies losses.
$BEA 0.2567 -11.36% Audiera, $171M turnover but heavy drop, capital fleeing decisively, panic spreading.
$OUST 45.94 -11.36% Ouster LiDAR US stock mapping, $2.02M turnover, stock logic highly linked to US market, extremely volatile.
💡 Afternoon Market Summary
1. Main themes collapse, defense dominates
Compared to midday, the 4:30 PM session shows not only no easing of declines but intensification. Average losses on the losers list far exceed gains on the winners list, market sentiment clearly deteriorated. Previous hotspots (AI, storage, Korea ETF) are hard hit.
2. Short ETFs dominate gainers
Top two gainers are SKDD (2x short Hynix) and SOXS (3x short semiconductor), clearly reflecting ongoing pullback in US tech sector (especially storage and chips). Capital is using crypto contracts to short US stock targets for hedging or speculation.
3. AI sector collapse across the board
ZHIPU (Zhipu) down another 14%, KAITO, WLD not in top ten but also falling. After prior frenzied hype, AI concept is in profit-taking stampede phase, any rebound may be a bull trap.
4. Danger signal: BEA (Audiera)
BEA fell 11.36% on $171M high volume, typical sign of major capital fleeing at any cost. Such stocks often continue to drift down, avoid catching falling knives.
5. Harsh reality of leveraged products
KORU (-17.86%) and SKUU (-13.56%) plunge alongside SKDD and SOXS surges vividly show the damage when direction is wrong on leveraged ETFs. Especially in choppy down markets, leverage decay significantly erodes principal.
📌 Strategy Tips
Market is clearly in risk-off mode. US tech sector adjustment pressure is transmitted to crypto contract market via short ETFs. In this environment:
· Abandon illusions: do not try to bottom-pick AI or storage related losers, trend is broken.
· Beware of bull traps: besides short ETFs, other gainers are mostly small-cap pulses or oversold rebounds, no main theme, chasing gains easily traps.
· Watch GPS: despite rise, $278M turnover fails to push price, typical distribution or high divergence signal, may fall further if market weakens.
Evening operations should focus on defense, strictly control positions, wait for clear direction after US market opens. In weak markets, cash is king.
#交易之声：你的经验值得被听到 #财报观察员: Xiaomi is about to release its earnings report, which business line do you favor more?
I'm Cige, Xiaomi will release its earnings report after the market closes tonight.
The market expects revenue of ¥108.8 billion, a year-on-year decline of about 6%, and adjusted net profit of about ¥6 billion. The three lines of smartphones, automobiles, and AIoT are advancing simultaneously.
In terms of smartphones, shipments in Q1 were 33.8 million units, down 19% year-on-year, but ASP rose 8.2% year-on-year to ¥1310, a record high. Volume down, price up, premiumization is being realized.
In automobiles, the SU7 series delivered 104,200 units in Q2, with a gross margin of 20.1%, and losses narrowed from ¥3.1 billion in Q1 to ¥2.06 billion. Scale effects are taking effect, and breakeven is not far off.
In AIoT, the 618 shopping festival drove Q2 IoT revenue to grow 28% quarter-on-quarter to ¥31.6 billion, with a clear recovery in major appliances and smart home.
The variable in Q3 is that if storage chip prices peak and fall, there is room for smartphone gross margin recovery. New automobile models will increase volume, and revenue contribution will continue to expand. Xiaomi's Q3 is more interesting than Q2.
Consumer electronics demand is recovering, AIoT is reviving, and the global tech hardware chain is emerging from the bottom. BTC, as the underlying asset of the computing power economy, is linked to the prosperity of tech hardware. When smartphones sell well, chip demand is stable, and capital expenditure on computing infrastructure will not stop.
That's all from Cige, savor it. $BTC $ETH $SNDK US Treasury yields fluctuate repeatedly, with BTC and ETH each moving independently; the underlying logic isn't that mysterious.
Many people like to generalize: US Treasury yields down = crypto market up, US Treasury yields up = crypto market down. But looking at the candlesticks repeatedly, divergences happen quite often.
$BTC is currently treated by large funds as a digital gold substitute, closely tracking real US Treasury yields and the US dollar index. When the yield curve jitters, institutional risk control models adjust parameters, and BTC is usually the first to be affected.
$ETH is different. It carries three burdens: selling pressure from staking unlocks, sluggish gas fee revenue, and a quiet ecosystem after Layer 2 traffic diversion. When the US Treasury environment warms up, BTC can hold steady, but ETH often struggles to rebound, with weak bounces and leading the declines.
The core point is: US Treasuries address the big-picture question of "whether to buy," but they don't solve the narrative problem of "why buy ETH."
Liquidity is just a ticket, not a reason for price increases. No matter how good the macro environment is, if on-chain revenue doesn't pick up and whales keep net outflows, ETH underperforming BTC is normal. Don't treat macro as a lifesaver; watching on-chain data is more effective than fixating on US Treasury yields.
#BTC #ETH #MacroAnalysis #USTreasuryYields #TradingAwareness#财报观察员：小米即将发布财报，你更看好哪条业务线？
Xiaomi's earnings report is coming, can Lei Jun take off this time?
I glanced at the market, hovering around 3.36, the market is waiting for the numbers to land.
The first quarter of 2026 is actually quite tough, storage chips have surged crazily, DRAM contract prices have cumulatively risen over 340%, and phone costs are unbearable. Xiaomi's Q1 revenue was 99.1 billion, down 10.9% year-on-year, but the phone ASP hit a historic high of 1310 yuan, indicating that the premiumization hasn't stopped.
The biggest variable is still cars. In Q1, 80,000 cars were sold, generating 19 billion in revenue, but operating losses were 3.1 billion. The conference call mentioned new cars in the second half, and the full-year target of 550,000 units remains unchanged. Lei Jun's targets are always aggressive; whether the volume can be pushed up in the second half is key.
The pressure on phones is also significant; the 18 series cut the Ultra, with Pro and Pro Max launching first in September. Storage price increases will last at least until the end of 2027, and costs can't be controlled. But from another perspective, Xiaomi is already using premiumization to hedge cost pressures; while others raise prices, it raises ASP, and this rhythm hasn't faltered.
#波动雷达：币种异动观察 ——$XIAOMI $SNDK, $MU, $WDC, and $STX rising together indicates that capital has started buying the "warehouse layer" of AI.
A very obvious phenomenon in the past two days is that U.S. stock storage and hard drive chains are strengthening together. $SNDK, $MU, $WDC, and $STX have all been brought up in market discussions, even companies related to hard drives and optical communications are being driven. This shows that AI trading has entered a diffusion stage: no longer just buying GPU and model companies, but starting to buy the entire data center’s warehouses, pipelines, and infrastructure.
AI systems are not just about computing power. Training and inference require GPUs, GPUs need HBM, model services require SSDs and NAND, massive data still needs HDDs, and data centers need networks, optical modules, power, and cooling. In the past, the market only focused on the most expensive parts; now it is starting to complete the entire industry chain map. Storage and hard drives are a rediscovered piece of this map.
The logic behind $SNDK and $MU leans toward high-performance storage and memory, while $WDC and $STX are more involved with enterprise storage, hard drives, and massive data. Many people think AI is all about high-speed memory and flash storage, no longer needing hard drives, but this is too simplistic. AI training data, videos, logs, backups, cold data, and enterprise data lakes cannot all be stored in the highest-cost storage. The larger the data volume, the more important tiered storage becomes. Hot data uses high-speed storage, cold data uses low-cost storage, making the entire system economically viable.
This is why $STX and $WDC can also be driven by AI. They are not the core high-end bottlenecks of AI, but they participate in the data explosion. The more AI spreads, the more data enterprises save, the more data models need to call, and the more complex the storage system becomes. Data centers don’t just buy the most expensive chips; they also need to buy enough hard drives and storage devices.
But this diffusion trading carries a risk: second-layer assets have high elasticity and also large drawdowns. Core leaders like $NVDA have platform moats, while storage and hard drive companies are more easily affected by cycles and order rhythms. If AI capital expenditure continues to rise, they will benefit; if cloud providers start controlling spending, the supply chain will be the first to have expectations cut.
So when writing about this sector, it’s not enough to say "AI storage will all rise." A more accurate statement is: after AI leaders’ valuations rise, capital begins to seek infrastructure segments that have not yet been fully priced. The commonality of $SNDK, $MU, $WDC, and $STX is not that they are AI leaders, but that they all sell indispensable data carrying capacity for the AI era.
The first layer of AI is computing power, the second layer is bandwidth, and the third layer is storage. The market is finally starting to realize: the smarter the model, the more important the underlying data warehouse becomes. $NVDA breaking below the key $221 level increases downside risk, with risk appetite in the tech sector clearly under pressure. Institutional heavyweights are significantly reducing positions; Dan Loeb liquidated 190,000 shares, Soros bought 400,000 put options defensively. Despite positive cooperation news, the stock price continues to fall, indicating signs of capital withdrawal. If the price remains below the $221.56 Bollinger middle band, selling pressure may further extend to the $215 support level. However, if physical AI ecosystem cooperation drives a volume breakout above the $225 resistance zone, the short-term bearish scenario will be invalidated.
#Anthropic年化营收达650亿美元 #财报观察员：小米即将发布财报，你更看好哪条业务线？$COMP 今天放量 14 倍暴涨 7.38%，创 60 日新高 18.53。催化剂很清楚：Compound 批了 5200 万美元创纪录预算，宣布正式从加密原生借贷，转向服务机构客户和现实世界资产（RWA）。 先泼盆冷水看转型动机。TVL 从峰值 120 亿美元跌到 12 亿，缩水九成——这不是"高瞻远瞩的战略升级"，是"老路走不动了"。加密原生借贷的用户和资金在流失，不转机构、不做 RWA，就是继续阴跌。所以这波转型，本质是求生。但赛道数据也真实：DeFi 里实际部署的 RWA 已经到 39.8 亿美元，一年涨了约 6 倍，私人信贷是主力——蛋糕在变大，Compound 想切的是这块。 第二个问号来自安全侧，而且时间点很微妙。就在 $COMP 放量创新高的同时，OpenZeppelin 的联创 Manuel Aráoz 公开发了条推：他觉得所有 DeFi 都不安全，已经建议亲友清仓包括 Aave、MakerDAO、Compound 在内的 DeFi 仓位。理由是 AI 编程代理找漏洞已经到了"超人水平"——攻击者只要找到那一个漏洞就够，防御者得把全部漏洞都修掉，攻防严重不对称。再叠#Gold stands above $4430, options capital turns bullish But I have to pour cold water first: don't go all in in excitement
Today spot gold once broke 4430, up 0.4% intraday, and New York futures gold even touched 4490. The key is not how much it rose, but the way it broke through — it retook the 50-day moving average and pierced the previous high resistance since mid-June, directly triggering a wave of programmatic short covering.
Options skew has really turned. The data is clear: gold options have shifted from buying put protection to buying calls. A guy spent $5.55 at once to buy 8,000 November 460 SPDR Gold Trust calls, and gold funds also recorded the strongest inflow since January. Institutions are not just bullish in words; they are betting real money.
But brothers, don't forget, the root behind this wave is still "de-dollarization." Central banks net purchased 288.9 tons of gold in Q2, a month-on-month surge of 411%, and our central bank has increased holdings for 21 consecutive months; the probability of a rate hike in September dropped to 30%, with both nonfarm payrolls and retail sales collapsing. If the US dollar credit really has another mishap, gold's safe-haven foundation is much stronger than $BTC, which is why I often say they are two sides of the same narrative.
My approach: continue holding spot positions, don't chase above 4430, consider adding on a pullback to 4400–4410; set stop loss at 4360. Those wanting to get in, don't chase the 460 calls, the premium has already been eaten, the cost-performance ratio is average.